Intel NetStructure® MPCBL0001
High Performance Single Board
Computer
Technical Product Specification
July 2005
Order Number: 273817-007
INFORMATION IN THIS DOCUMENT IS PROVIDED IN CONNECTION WITH INTEL® PRODUCTS. NO LICENSE, EXPRESS OR IMPLIED, BY
ESTOPPEL OR OTHERWISE, TO ANY INTELLECTUAL PROPERTY RIGHTS IS GRANTED BY THIS DOCUMENT. EXCEPT AS PROVIDED IN
INTEL'S TERMS AND CONDITIONS OF SALE FOR SUCH PRODUCTS, INTEL ASSUMES NO LIABILITY WHATSOEVER, AND INTEL DISCLAIMS
ANY EXPRESS OR IMPLIED WARRANTY, RELATING TO SALE AND/OR USE OF INTEL PRODUCTS INCLUDING LIABILITY OR WARRANTIES
RELATING TO FITNESS FOR A PARTICULAR PURPOSE, MERCHANTABILITY, OR INFRINGEMENT OF ANY PATENT, COPYRIGHT OR OTHER
INTELLECTUAL PROPERTY RIGHT.
Intel products are not intended for use in medical, life saving, life sustaining, critical control or safety systems, or in nuclear facility applications. Intel
may make changes to specifications and product descriptions at any time, without notice.
Designers must not rely on the absence or characteristics of any features or instructions marked “reserved” or “undefined.” Intel reserves these for
future definition and shall have no responsibility whatsoever for conflicts or incompatibilities arising from future changes to them.
The Intel NetStructure
cause the product to deviate from published specifications. Current characterized errata are available on request.
This document and the software described in it are furnished under license and may only be used or copied in accordance with the terms of the
license. The information in this document is furnished for informational use only, is subject to change without notice, and should not be construed as a
commitment by Intel Corporation. Intel Corporation assumes no responsibility or liability for any errors or inaccuracies that may appear in this
document or any software that may be provided in association with this document. Except as permitted by such license, no part of this document may
be reproduced, stored in a retrieval system, or transmitted in any form or by any means without the express written consent of Intel Corporation.
Contact your local Intel sales office or your distributor to obtain the latest specifications and before placing your product order.
Copies of documents which have an ordering number and are referenced in this document, or other Intel literature may be obtained by calling
1-800-548-4725 or by visiting Intel's website at http://www.intel.com.
AlertVIEW, AnyPoint, AppChoice, BoardWatch, BunnyPeople, CablePort, Celeron, Chips, CT Connect, CT Media, Dialogic, DM3, EtherExpress,
ETOX, FlashFile, i386, i486, i960, iCOMP, InstantIP, Intel, Intel logo, Intel386, Intel486, Intel740, IntelDX2, IntelDX4, IntelSX2, Intel Create & Share,
Intel GigaBlade, Intel InBusiness, Intel Inside, Intel Inside logo, Intel NetBurst, Intel NetMerge, Intel NetStructure, Intel Play, Intel Play logo, Intel
SingleDriver, Intel SpeedStep, Intel StrataFlash, Intel TeamStation, Intel Xeon, Intel XScale, IPLink, Itanium, LANDesk, LanRover, MCS, MMX, MMX
logo, Optimizer logo, OverDrive, Paragon, PC Dads, PC Parents, PDCharm, Pentium, Pentium II Xeon, Pentium III Xeon, Performance at Your
Command, RemoteExpress, Shiva, SmartDie, Solutions960, Sound Mark, StorageExpress, The Computer Inside., The Journey Inside,
TokenExpress, Trillium, VoiceBrick, Vtune, and Xircom are trademarks or registered trademarks of Intel Corporation or its subsidiaries in the United
States and other countries.
† Hyper Threading Technology (HT Technology) requires a computer system with an Intel® Pentium® 4 processor supporting HT Technology and an
HT Technology-enabled chipset, BIOS and operating system. Performance will vary depending on the specific hardware and software you use. See
http://www.intel.com/info/hyperthreading/ for more information including details on which processors support HT Technology.
*Other names and brands may be claimed as the property of others.
July 2005007Added Table 7. Modified tables 3, 9, 13, 14, and 53; Fig. 21; and Section 10.5.
April 2005006New text in sections 3.2.9, 6.5, 10.3.1, and tables 2, 3, and 6.
February 2005005New text, figures; added Section 18, “Agency Information—Class B”.
November 2004 004
June 2004003SRA Release - changed from release 002 to current.
January 2004002Pre-SRA Release.
October 2003001Initial public release of this document
10Intel NetStructure
Changes to figures 12, 13; changes to table 2, 3, 48, 77 and 81; added example
to Section 3.2.5.
®
MPCBL0001 High Performance Single Board Computer
Technical Product Specification
Introduction
Introduction1
1.1Document Organization
This document gives technical specifications related to the Intel NetStructure® MPCBL0001 High
Performance Single Board Computer. The MPCBL0001 is designed following the standards of the
Advanced Telecommunications Compute Architecture (AdvancedTCA*) Design Guide for high
availability, switched network computing. This document is intended for support during system
product development and while sustaining a product. It specifies the architecture, design
requirements, external requirements, board functionality, and design limitations of the
MPCBL0001 Single Board Computer.
The following summarizes the focus of each chapter in this document.
Chapter 1, “Introduction” gives an overview of the information contained in the Intel
NetStructure
Specification as well as a glossary of acronyms and important terms.
Chapter 2, “Features Overview” introduces the key features of the MPCBL0001. It includes a
functional block diagram and a brief description of each block.
Chapter 3, “Hardware Management Overview”provides a high-level overview related to IPMI
implementation based on PICMG* 3.0 and IPMI v1.5 specifications in the MPCBL0001 SBC.
Chapter 4, “Connectors” includes an illustration of connector locations, connector descriptions,
and pinout tables.
Chapter 5, “Addressing” summarizes the information you need to configure the MPCBL0001.
Included are the PCI configuration map, Configuration Address register, Configuration Data
register, I/O address assignments, memory map, and IPMC addresses.
Chapter 6, “Specifications” contains the mechanical, environmental, and reliability specifications
for the MPCBL0001.
Chapter 7, “BIOS Features” provides an introduction to the Intel/AMI BIOS, and the System
Management BIOS, stored in flash memory on the MPCBL0001.
Chapter 8, “BIOS Setup” describes the interactive menu system of the BIOS Setup program. The
menu allows a user to configure the BIOS for a given system.
Chapter 9, “Error Messages” lists BIOS error messages, Port 80h POST codes, and bus
initialization checkpoints, and provides a brief description of each.
®
MPCBL0001 High Performance Single Board Computer Technical Product
Chapter 10, “Operating the Unit” provides specifics for configuring the MPCBL0001, including
BIOS configuration and jumper settings.
Chapter 11, “Maintenance” includes supervision and diagnostics information.
Chapter 13, “Component Technology” lists the major components used on the MPCBL0001.
Intel NetStructure
Technical Product Specification
®
MPCBL0001 High Performance Single Board Computer 11
Introduction
Chapter 14, “Warranty Information” provides warranty information for Intel® NetStructureTM
products.
Chapter 15, “Customer Support” provides informat ion on how to contact customer support.
Chapter 16, “Certifications” and Chapter 17, “Agency Information—Class A” document the
regulatory requirements the MPCBL0001 is designed to meet.
Appendix A, “Reference Documents” provides a list of data sheets, standards, and specifications
for the technology designed into the MPCBL0001.
Appendix B, “List of Supported Commands (IPMI v1.5 and PICMG 3.0)”provides lists of
commands supported by IPMI v1.5 and PICMG Specification 3.0.
1.2Glossary
For ease of use, numeric entries are listed first with alpha entries following. Acronyms and terms
are then entered in their respective place.
ACPIAdvanced Configuration and Power Interface.
AdvancedTCAAdvanced Telecommunications Compute Architecture
BIOSBasic Input/Output Subsystem. ROM code that initializes the computer
and performs some basic functions.
BladeAn assembled PCB card that plugs into a chassis.
DIMMDual Inline Memory Module. Small card with memory on it used for
MPCBL0001.
DMIDesktop Management Interface
EEPROMElectrically Erasable Programmable Read-Only Memory
Fabric BoardA board capable of moving packet data between Node Boards via the
ports of the backplane. This is sometimes referred to as a switch.
Fabric SlotA slot supporting a link port connection to/from each Node Slot and/or
out of the chassis.
Hyper-Threading Technology
†
HT T echnology allows a single (or dual) physical processor, to appear as
two (or quad) logical processors to a HT Technology-aware operating
system.
2
I
C*Inter-IC [Integrated Circuit]. 2-wire interface commonly used to carry
management data.
IBAIntel
®
Boot Agent. The Intel Boot Agent is a software product that
allows your networked client computer to boot using a program code
image supplied by a remote server.
IDEIntegrated Device Electronics. Common, low-cost disk interface.
IPMBIntelligent Platform Management Bus. Physical 2-wire medium to carry
IPMI.
12Intel NetStructure
®
MPCBL0001 High Performance Single Board Computer
Technical Product Specification
Introduction
IPMCIntelligent Platform Management Controller. ASIC in baseboard
responsible for low-level system management.
IPMIIntelligent Platform Management Interface. Programming model for
system management.
KCSKeyboard Controller Style interface.
LPC BusLos Pin Count Bus. Legacy I/O bus that replaces ISA and X-bus. See the
Low Pin Count (LPC) Interface Specification.
MTBFMean Time Between Failure. A reliability measure based on the
probability of failure.
NEBSNational Equipment Building Standards. Telco standards for equipment
emissions, thermal, shock, contaminants, and fire suppression
requirements.
NMINon-Maskable Interrupt. Low-level PC interrupt.
Node BoardA board capable of providing and/or receiving packet data to/from a
Fabric Board via the ports of the networks. The term is used
interchangeably with SBC.
MPCBL0001Single or dual processor Single Board Computer with Fibre Channel.
MPCBL0002Single or dual processor Single Board Computer without Fibre Channel.
Node SlotA slot supporting port connections to/from Fabric Slot(s). A Node slot is
intended to accept a Node Board
Physical PortA port that physically exists. It is supported by one of many physical
(PHY) type components.
PMCPCI Mezzanine Card. IEEE1386 standard for embedded PCI cards. They
mount parallel to the SBC.
ROMRead-Only Memory.
SBCSingle Board Computer. This term is used interchangeably with Node
Board.
SELSystem Event Log. Action logged by management controller.
SFPSmall Form Factor Pluggable receptacle for the front panel Fibre
Channel interfaces.
SMBusSystem Management Bus. Similar to I
2
C
SMISystem Management Interrupt. Low-level PC interrupt which can be
initiated by chipset or management controller. Used to service IPMC or
handle things like memory errors.
SMS, SMSCStandard Microsystems Corporation*
USBUniversal Serial Bus. General-purpose peripheral interconnect,
operating at 1-12 Mbps.
Intel NetStructure
Technical Product Specification
®
MPCBL0001 High Performance Single Board Computer 13
Features Overview
Features Overview2
2.1Application
The Advanced T elecommunications Compute Architecture (AdvancedTCA) standards define open
architecture modular computing components for carrier-grade, communications network
infrastructure. The goals of the standards are to enable blade-based modular platforms to be:
• cost effective
• high-density
• high-availability
• scalable
These systems use a fabric I/O network for connecting multiple, independent processor boards, I/O
nodes (e.g., line cards), and I/O devices (e.g., storage subsystem).
The MPCBL0001 SBC is designed per the AdvancedTCA Design Guide for High Availability,
Switched Network Computing. Bulk storage for the system is connected through optional dual
Fibre Channel interfaces. The MPCBL0001FXX SBC includes a Fibre Channel controller. The
MPCBL0001NXX SBC does not have the Fibre Channel controller.
2.2Functional Description
This topic defines the architecture of the MPCBL0001 SBC through descriptions of functional
blocks. Figure 1, “Intel NetStructure® MPCBL0001 SBC Block Diagram” on page 15 shows the
functional blocks of the MPCBL0001 SBC. The MPCBL0001 SBC is a dual processor, hotswappable SBC with backplane connections to dual Gigabit Ethernet star networks and dual Fibre
Channel star arbitrated loops.
The SBC incorporates an Intelligent Platform Management Contro ller that monitors critical
functions of the board, responds to commands from the shelf manager, and reports events.
Power is supplied to the MPCBL0001 SBC through two redundant -48 V power supply
connections. Power for on-board hardware management circuitry is provided through a standby
converter on the power mezzanine. This converter, along with all the other converters on the power
mezzanine are fed by the diode OR'd -48 V supply from the backplane.
The SBC has provision for the addition of a PMC device and supports 32-bit and 64-bit transfers at
33 MHz and 66 MHz. The SBC also offers one USB and one service terminal interface. An
overview of each block follows.
MPCBL0001 High Performance Single Board Computer 15
Technical Product Specification
Low Voltage
Low Voltage
Intel® Xeon™
Intel® Xeon™
Processor
Processor
Low Voltage
Low Voltage
Intel® Xeon™
Intel® Xeon™
Processor
Processor
Dual Fibre Channel Ports to Fabric Interface
Dual Gigabit Ethernet Ports to Base Interface
Features Overview
2.2.1Low Voltage Intel® Xeon
(U36)
The MPCBL0001 SBC supports up to two Low Voltage Intel® Xeon™ processors (see Figure 20,
“Intel NetStructure® MPCBL0001 Component Layout” on page 90 for locations). The Low
Voltage Xeon processor incorporates Intel
Front-Side Bus, allowing performance levels that are significantly higher than previous generations
of IA-32 family processors. The processors include the following features:
• Intel & OEM EEPROM and thermal sensor manageability features
• Supports single and dual processor configurations
• Throttling enabled for protection against high temperatures
™
Processor CPU-0 (U35), CPU-1
®
NetBurst™ microarchitecture and a high-bandwidth
The Low Voltage Xeon processor host bus utilizes a split-transaction, deferred-reply protocol. The
host bus uses source-synchronous transfer of address and data to improve throughput at the 100 or
133 MHz bus frequency (depending on processor model). Addresses are transferred at 2X the bus
frequency while data is transferred at 4X the bus frequency, resulting in peak data transfer rates up
to 3.2 or 4.3 GBytes/s.
In addition to the NetBurst microarchitecture, the Low Voltage Intel Xeon processor includes a
groundbreaking technology called Hyper-Threading Technology
Technology improves processor performance for multithreaded applications or multitasking
environments by supporting multiple software threads on each processor.
Low Voltage Intel Xeon processors require their package case temperatures to be operated below
an absolute maximum specification. If the chassis ambient temperature exceeds a level whereby
the processor thermal cooling subsystem can no longer maintain the specified case temperature, the
processors will automatically enter a mode called Thermal Monitor to reduce their case
temperatures. Thermal Monitor controls the processor temperature by modulating the internal
processor core clocks, thereby reducing internal power dissipation, and does not require any
interaction by the Operating System or Application. Once the case temperatures have reached a
safe operating level, the processor will return to its non-modulated operating frequency. See the
Low Voltage Intel Xeon processor datasheet, referenced in Appendix A, “Reference Documents”,
for further details.
An optional ITP700 port connection is included to facilitate debug and BIOS/software
development efforts. This JTAG connection to the processors utilizes voltage-signaling levels that
are specific to the Low Voltage Xeon processor family. These levels must not be exceeded or
processor damage may occur. Please refer to Intel document ITP700 Debug Port Design Guide,
order number 249679-005 for additional information on the ITP connector pin defi nitions.
†
(HT Te chnology). HT
16Intel NetStructure
®
MPCBL0001 High Performance Single Board Computer
Technical Product Specification
2.2.2Chipset
The Intel® E7501 chipset consists of three major components:
®
• Intel
• Intel
• Intel
See Figure 20, “Intel NetStructure® MPCBL0001 Component Layout” on page 90 for their
locations.
2.2.2.1Intel® E7501 Memory Controller Hub (U22)
The Intel® E7501 Memory Controller Hub (MCH) interfaces between the processor system bus
and the memory and I/O subsystems.
Significant features are listed below:
• System/host bus features:
E7501 Memory Controller Hub (MCH)
®
82801CA I/O Controller Hub 3 (ICH3)
®
82870P2 64-bit PCI/PCI-X Controller Hub 2 (P64H2)
— Supports dual processors at either 400 or 533 MT/s or a bandwidth of 3.2 or 4.3 GBytes/s
— Supports a 36-bit system bus addressing model
— 12 deep in-order queue, two deep defer queue
Features Overview
Note: The current MPCBL0001 is designed to run with the Intel
processor frequency, the processor side bus (PSB) will run at 400 MT/s with a bandwidth of 3.2
GBytes/s.
• Memory subsystem features:
— 144-bit wide (72-bit x 2), DDR-266 memory interfaces with 3.2 or 4.3 GByte/s bandwidth
— Supports x72, registered DDR-266 ECC DIMMs using 64-, 128-, 256-, and 512-Mbit
SDRAMs
— Supports a maximum of 16 GBytes of memory (MPCBL0001 SBC implementation
supports a maximum of 8 Gbytes).
— Supports S4EC/D4ED ChipKill* ECC (x4 ChipKill)
• Corrects all bit errors within a single 4-bit nibble
• Detects all errors contained within two 4-bit nibbles
• Memory scrubbing supported
— Supports up to 32 simultaneous open pages
— Hardware support for auto-initialization of memory with valid ECC
• I/O features:
— Hub interface A provides HI 1.5 connection for ICH3
• 266 MB/s data bandwidth with parity protection
• 8 bits wide, 66 MHz clock, 4x data transfer (quad-pumped)
• Supports 64-bit inbound addressing, 32-bit outbound addressing
— Hub interfaces B and C provide HI2.0 connections for two P64H2s
• 1 GByte/s data bandwidth with ECC protection in each direction
®
LV Xeon® 2.0 GHz processor. At this
Intel NetStructure
Technical Product Specification
®
MPCBL0001 High Performance Single Board Computer 17
Features Overview
• 16-bits wide, 66 MHz clock, 8x data transfer (octal pumped)
The MCH I/O subsystems interface incorporates four hub interfaces. Each Hub interface is a pointto-point connection between the MCH and an I/O bridge/device. The various components of the
chipset communicate via these connected hub interfaces:
• The first hub link connects the MCH to the ICH3.
• The next two hub link interfaces connect the MCH to P64H2 components.
• The remaining hub link is unused.
2.2.2.2Intel® 82801CA I/O Controller Hub 3 (U7)
The Intel® 82801CA I/O Controller Hub 3 (IHC3) provides the legacy I/O subsystem and
integrates advanced I/O functions. ICH3 features are listed below:
• IDE interface controller
• Three Universal Host Controller Interface (UHCI)
• USB host controllers supporting up to 6 ports (MPCBL0001 SBC implementation supports
one port on the front panel)
• Integrated I/O APIC
• SMBus 2.0 controller
• LPC interface
• Watchdog timer #3 (see “Watchdog Timers (WDTs)” on page 62)
• PCI 2.2 bus interface supporting 32bit/33 MHz operation
• Connects to MCH through Hub Interface A (HI 1.5)
The MPCBL0001 SBC implements one USB port and does not use the ICH3 PCI connection.
2.2.2.2.1PCI Bus Master IDE Interface (J24)
The ICH3 acts as a PCI based, enhanced IDE, 32-bit interface controller for intelligent disk drives
that have disk controller electronics onboard. The SBC includes a single 40-pin (2 x 20) IDE
connector (J24) that supports one master or one slave device. See Figure 20, “Intel NetStructure®
MPCBL0001 Component Layout” on page 90 drawing for its location. The IDE controller
provides support for an internally mounted 2.5” hard disk. The IDE control ler has the following
features:
• PIO and DMA transfer modes
• Mode 4 timings
• Supports Ultra ATA33/66/100 synchronous DMA
• Buffering for PCI/IDE burst transfers
• Master/slave IDE mode
• Support for up to two devices (Master/Slave) via a single primary IDE connector
(MPCBL0001 SBC implementation supports one optional physical 2.5" IDE device)
Note: Incorporating an optional IDE Hard Disk drive may significantly impact the Reliability
Specifications in Section 6.3.
18Intel NetStructure
®
MPCBL0001 High Performance Single Board Computer
Technical Product Specification
Features Overview
Note: Performance of the IDE interface may be impacted by the DMA mode and type of DMA transfers
used. Even though the BIOS automatically sets the DMA mode/type, the OS could downgrade the
DMA transfer mode. Check the operating system documentation to see what DMA mode is used
by default and whether it is possible to change to a higher performance DMA mode.
The two P64H2 devices provide the system’s high-performance PCI bus support. See Figure 20,
“Intel NetStructure® MPCBL0001 Component Layout” on page 90 for their locations. Each
P64H2 component supports two independent, 64-bit, PCI/PCI-X interfaces. 32-bit/33 MHz and 64bit/66 MHz PCI bus modes are also supported. Each PCI bus interface features:
• PCI-X 1.0 Specification compliance
• PCI Specification 2.2 compliance
• PCI-PCI Bridge Rev 1.1 compliance
• PCI Hot Plug 1.0 compliance
• I/O APIC supporting up to 24 interrupts (16 external pins)
• PCI peer-to-peer write capability between PCI ports
• SMBus target for Out-of-Band access to all internal PCI registers
Each of the two P64H2 devices (U14, U24) included on the MPCBL0001 SBC provides the bridge
to two independent PCI bus connections, as shown in Table 1, “P64H2 Interfaces” on page 19.
Table 1. P64H2 Interfaces
P64H2 Device Interface
U24PCI-X interface to the optional dual Fibre Channel controller
U14
The two high-speed communications interfaces (Gigabit Ethernet and Fibre Channel) are located in
separate P64H2 devices to maximize data throughput. A single HI-2 hub link connection from the
P64H2 to the MCH provides a >1 Gbyte/s bandw idth back to memory and the processor System
Bus.
• PCI-X interface to the dual Gigabit Ethernet controller
• 64-bit/66 MHz PCI bus for a plug-in PMC card
2.2.3Memory (J8, J9, J10, J11)
Four DDR 266 DIMM sockets make up the memory subsystem. See Figure 20, “Intel
NetStructure® MPCBL0001 Component Layout” on page 90 for their locations. The MCH defines
two memory channels operating in parallel to logically create a 144-bit wide memory data path.
ECC is generated and checked across 128 bits of data, allowing for significant improvement in
error correction.
Due to this architecture, DDR DIMMs must be installed in matched pairs. Memory DIMM
configurations ranging from 512 MBytes to 8 GBytes in 512 MByte increments are supported.
Intel NetStructure
Technical Product Specification
®
MPCBL0001 High Performance Single Board Computer 19
Features Overview
2.2.3.1Memory Ordering Rule for the MCH
Platforms based on the E7501 chipset require DDR DIMMs to be populated in matched pairs in a
specific order. Start with the two DIMMs furthest from the MCH in a “fill-farthest” approach (see
Figure 2). This requirement is based on the signal integrity requirements of the DDR interface.
Figure 2. Memory Ordering
MCH, U22
J8
Fill
Last
J9
Fill
First
J11
J10
2.2.4I/O
2.2.4.1Super I/O (U28)
The Super I/O device (SIO) is an SMSC LPC47B272 enhanced Super I/O controller. The SIO
connects to the ICH3 through its LPC bus connection. The SIO provides support for the front panel
serial port (J17, see page 70). There is no front-panel connection to the legacy keyboard and mouse
PS/2 ports. Keyboard and mouse support are provided by the USB connection (J12, see page 77).
See Figure 13 for connector locations.
20Intel NetStructure
B0894-01
®
MPCBL0001 High Performance Single Board Computer
Technical Product Specification
To facilitate debug and BIOS development, SIO connections such as legacy (PS/2) keyboard/
mouse and floppy may be provided on initial board revisions. Software must not rely on the
presence of these connections on future board revisions.
2.2.4.2Real-Time Clock
The MPCBL0001 SBC real-time clock is integrated into the ICH3. It is derived from a 32.768 KHz
crystal with the following specifications:
• Frequency tolerance @ 25 ºC: ±20ppm
• Frequency stability: maximum of -0.04ppm/(ΔºC)
• Aging ΔF/f (1
st
year @ 25 ºC): ±3ppm
• ±20ppm from 0-55 ºC and aging 1ppm/year
The real-time clock is powered by a 0.22F SuperCap* capacitor when main power is not applied to
the board. This capacitor powers the real-time clock for a minimum of two hours while external
power is removed from the MPCBL0001 SBC.
See Section 3.13, “Watchdog Timers (WDTs)” on page 62 for information about the real-time
clock timers.
Features Overview
2
2.2.4.3Timer0 Capabilities
Timer0, integrated inside the ICH3, is an 8254 compatible timer. This timer is set up to generate a
periodic waveform that creates the edge for the timer0 interrupt. The interrupt is received by the
ICH3 APIC and communicated to the CPU(s).
MPCBL0001 provides a high-precision 14.318 MHz crystal clock source as the reference for the
timer0 counters. To improve timing accuracy, the crystal used is a low-PPM, high-stability
component with the following specifications:
• Frequency tolerance (25º C): ±10ppm
• Temperature characteristics (-10º C to +60º C): ±5ppm
• Aging: ±1ppm per year max
This timer does not operate when board power is removed.
2.2.4.4Gigabit Ethernet (U13)
The MPCBL0001 SBC implements two Gigabit Ethernet interfaces, each of which is routed to the
fabric/switch slot through the backplane (J23, see page 75). There are no direct, external Ethernet
ports included on the SBC board. Each Ethernet connection utilizes an 82546 Dual Gigabit
Ethernet Controller, allowing support for 1000Mbits/s, 100Mbits/s and 10Mbits/s data rates.
The 82546 controller is optimized for designs using the PCI and the emerging PCI-X bus interface
extension. The MPCBL0001 SBC has a 133 MHz PCI-X bus connection. The integrated physical
layer circuitry (PHY) provides an IEEE 802.3 Ethernet Interface for 1000Base-T, 100Base-TX,
and 10Base-T applications.
MPCBL0001 High Performance Single Board Computer 21
Features Overview
• Host interface also compliant with the PCI-X addendum, Rev 1.0a, from 50 to 133 MHz
• Supports 64-bit addressing
• Efficient PCI bus master operation, supported by optimized internal DMA controller
• Supports advanced PCI commands such as MWI, MRM, and MRL, and PCI-X commands
such as MRD, MRB, and MWB
• Full IEEE 802.3ab auto-negotiation of speed, duplex, and flow-control configuration
• Complete full duplex and half duplex support
• Automatic MDI crossover operation for 100Base-TX and 10Base-T modes
• Automatic polarity correction
• Digital implementation of adaptive equalizer and canceller for echo and crosstalk
2.2.4.5Fibre Channel* (U23) - Optional
The QLogic* ISP2312 dual Fibre Channel controller is used for access to high-speed storage
subsystems. It is routed through backplane connector P23.
See Figure 20, “Intel NetStructure® MPCBL0001 Component Layout” on page 90 for its location.
This controller supports PCI and PCI-X bus interfaces. Burst mode master DMA transfers are
utilized for efficient usage of bus bandwidth during data transfers, and 8, 16, and 32-bit accesses
are supported as a PCI target. The controller appears as two independent Fibre Channel ports. A
PCI function is assigned to each port in the device’s PCI configuration space. Functions 0 and 1 are
used to configure FC ports 1 and 2, respectively.
— DMA channels (transmit, receive, command, auto-request, and auto-response)
• Support for JTAG boundary scan.
• Supports IP as well as other protocols; however there are currently no plans to validate
protocols other than SCSI_FCP.
Each Fibre Channel interface of the ISP2312 includes its own internal 16-bit RISC processor and
external 7.5 ns synchronous SRAM memory for instruction code and data. Parity protection is
provided on accesses to this memory. The SBC utilizes two 256 KByte (128Kx18) SRAMs, one for
each port, for the ISP2312 memory requirements.
An external 256 x 16 non-volatile EEPROM is used to store system configuration parameters and
PCI subsystem and subsystem vendor IDs. The first 128 bytes are used fo r function 0 parameters
and the second 128 bytes are used for function 1.
2.2.5PMC Connector (J25, J26, J27)
The MPCBL0001 SBC supports one 64-bit, 66 MHz PMC slot. The PMC slot is connected to the
second of two P64H2 hub controllers via PMC Connectors J25-J27. The PMC slot has an opening
in the front panel of the SBC that exposes the I/O connectors of the add-in PMC card. PMC cards
can only be added to or removed from this slot when the board is outside the system chassis. See
Figure 20, “Intel NetStructure® MPCBL0001 Compo nent Layou t” on page 90 for its location.
Features Overview
The PCI bus specification provides the means for backward compatibility with slower PMC cards
(32-bit or 33 MHz) through the use of the M66EN pin. A PMC card that does not support 66 MHz
operation grounds the M66EN pin when installed to inform the SBC hardware to provide a 33
MHz clock to this interface. Support for 32-bit only PMC cards is accomplished through the use of
the REQ64#/ACK64# PCI bus protocol.
The PMC slot provided by the SBC connects the PCI VI/O voltage pins to +3.3 V. This requires
use of PMC plug-in cards that support +3.3 V I/O signal levels. Only PM C plug-in cards
designated “+3.3 V only” or “univ e rsal” voltage I/O are supported. The PMC plug-in location
provides a key pin to prevent insertion of cards that do not meet this requirement. Note that +5 V
power is still supplied to the PMC pins designated for +5 V connections. The PMC is allotted 1.5 A
of current.
2.2.6Firmware Hub (U30, U33)
The MPCBL0001 SBC supports two 8Mbit (1 MByte) BIOS flash ROMs:
• Primary BIOS flash ROM (FWH0)
• Recovery BIOS flash ROM (FWH1)
The flash is allocated for BIOS and Firmware usage.
The SBC boots from the primary flash ROM under normal circumstances. During the boot process,
if the BIOS (or IPMC) determines that the contents of the primary flash ROM are corrupted, a
hardware mechanism is available to change the flash device select logic to the recovery flash
ROM. See Section 2.2.6.3, “Flash ROM Backup Mechanism” on page 24 for more information.
Each flash component has a separately write-protected boot block that prevents erasure when the
device is upgraded.
Intel NetStructure
Technical Product Specification
®
MPCBL0001 High Performance Single Board Computer 23
Features Overview
Flash ROM BIOS updates can be performed by an end user or a network administrator over the
LAN. The system should complete booting to an OS, MS-DOS* or logon to Linux* as root user.
The system should have a local copy of the flash program and the BIOS data files or have the
capability to copy the flash program and BIOS data files onto a local drive via the network. The
flash program has a command line interface to specify the path and the file name of the BIOS data
files. After completing the BIOS ROM update the user should shutdown and reset the system to let
the new BIOS ROM take effect. See Section 7.7, “BIOS Updates” on page 102 for more
information.
2.2.6.1FWH 0 (Main BIOS)
BIOS execute code off this flash and perform checksum validation of its operational code. This
checksum occurs in the boot block of the BIOS. The BIOS image is also stored in FWH0. When
user performs BIOS update, the BIOS image will be stored in FWH 0 only. FWH0 will also store
the factory default CMOS settings user configured CMOS settings.
1. When user "Load optimal defaults" from the BIOS setup screen, it restores the factory default
by copying the "Factory Default" settings from FWH0 to ICH3 (CMOS).
2. When user "Save custom defaults" from the BIOS setup screen, the changes will be made to
the CMOS settings on ICH3 and then copied from ICH3 to FWH0.
3. When user "Load custom defaults" from the BIOS setup screen, the "custom" CMOS settings
are copied from FWH0 to ICH3.
2.2.6.2FWH 1 (Backup/Recovery BIOS)
FWH 1 stores the recovery BIOS. In the event of checksum failure on the Main BIOS operational
code, BIOS will request BMC to switch FWH, so that the board will be able to boot up from FWH1
for recovery.
User is able to boot up the board from FWH1 by executing an OEM IPMI command as well (see
Section 3.7.1, “Reset BIOS Flash Type” on page 47 ).
2.2.6.3Flash ROM Backup Mechanism
The on-board Intelligent Platform Management Controller (IPMC) manages which of the two
BIOS flash ROMs is used during the boot process. The IPMC mon itors the boot progress and can
change the flash ROM selection and reset the processor.
The default state of this control configures the primary Firmware Hub (FWH) ROM device ID to
be the boot device; the secondary FWH is assigned the next ID. The secondary FWH responds to
the address range just below the primary FWH ROM in high memory.
The Intelligent Platform Management Controller sets the ID for both FWH devices. Boot accesses
are directed to the FWH with ID = 0; unconnected ID pins are pulled low by the FWH device. In
this way the IPMC may select which flash ROM is used for the boot process.
Refer to Section 3.7.1, “Reset BIOS Flash Type” on page 47 for a description of how to do this
manually.
24Intel NetStructure
®
MPCBL0001 High Performance Single Board Computer
Technical Product Specification
2.2.7Onboard Power Supplies
The main power supply rails on the MPCBL0001 SBC are powered from dual-redundant -48 V
power supply inputs from the backplane power connector (P10). There are also dual redundant,
limited current, make-last-break-first (MLBF) power connections. See Figure 20, “Intel
NetStructure® MPCBL0001 Component Layout” on page 90 for their location.
2.2.7.1Power Feed Fuses
As required by the PICMG 3.0 Specification, the MPCBL0001 SBC provides fuses on each of the
-48V power feeds and on the RTN connections as well . The fuses on the return feeds are critical to
prevent overcurrent situations if an ORing diode in the return path fails and there is a voltage
potential difference between the A and B return paths.
2.2.7.2ORing Diodes and Circuit Breaker Protection
The two -48 V power connectors are OR’d together. A current limiting FET switch is connected
between the OR’d -48 V and the primary DC-DC converters. The FET switch provides three
functions:
• A mechanism to electrically connect/disconnect the SBC to/from the two -48 V inputs.
• A soft-on function.
Features Overview
• An over-current circuit breaker feature.
2.2.7.3-48 V to +12 V Converter
This converter provides DC isolation between the -48 V and -48 V return connections and all of the
derived DC power on the MPCBL0001 SBC. Its output is connected to the SBC’s +12 V power
rail. The converter supplies a maximum of 9 A of current. The converter is enabled/disabled by the
onboard IPMC.
2.2.7.4-48 V to +5 V/+3.3 V Converter
This converter provides DC isolation between the -48 V and -48 V return connections and all of the
derived DC power on the MPCBL0001 SBC. Its output is connected to the SBC’s +5 V and 3.3 V
power rails. The converter supplies a maximum of 9 A of +5 V current and 9 A of +3.3 V current.
The converter is enabled/disabled by the onboard IPMC.
2.2.7.5Processor Voltage Regulator Module (VRM)
The Voltage Regulator Module (VRM) provides core power to the two Low Voltage Xeon
processors. The input to the VRM is connected to the +12 V power rail.
See Figure 20, “Intel NetStructure® MPCBL0001 Component Layout” on page 90 for its location.
Intel NetStructure
Technical Product Specification
®
MPCBL0001 High Performance Single Board Computer 25
Features Overview
The VRM controller is designed to support multiple processor core voltages selected by the voltage
identification (VID) pins on the processor. Logic provided on the SBC ensures that the VRM is not
enabled if the two processors request different VID codes. In addition, the VRM is disabled until
all other voltage converters indicate “power good.” The voltage regulator module is designed to
support up to two 43 W (TDP - Thermal Design Power) processors.
Note: The +5 VSB power rail only needs to supply at least 4.0 V to properly power any circuitry that uses
the +5 VSB rail when the payload power (i.e., processors, ethernet controller, etc.) is not turned on.
Any alerts from the +5 VSB sensor when the system is not in the M4, M5, or M6 states should be
ignored.
2.2.7.6IPMB Standby Power
This converter provides DC isolation between the -48 V and -48 V return connections and all of the
derived DC power on the MPCBL0001. Its output is connected to the IPMB and standby +5 V
power rail of the SBC. The converter supplies a maximum of 1.5 A of +5 V current. A +3.3 V
management voltage is derived from the IPMB power by means of a linear regulator circuit and is
used to power most of the IPMC functions. Standby power is derived from the -48 V rails and is
always available on the SBC unless the overall system power rail (-48 V) is shut down.
26Intel NetStructure
®
MPCBL0001 High Performance Single Board Computer
Technical Product Specification
Hardware Management Overview
Hardware Management Overview3
The Intelligent Platform Management Controller (IPMC) is an Intel-designed baseboard
management controller device manufactured by Philips Semiconductor* for Intel.
The high-level architecture of the baseboard management for MPCBL0001 is represented in the
block diagram below.
Figure 3. Hardware Management Block Diagram
ADM 1026
Wat chdog Tim er
CPU
(Low Voltage Int el
Xeon™)
Intel® E7501 Memory
Controller H ub (MCH)
ICH3
Intelligent Platform
Management Controller
(IPMC)
Flash
Mem ory
SRAM
®
NOT E:
IPMB A
I
P
M
B
B
I2C Bus
KCS interf ace
Logic C onnect ion
Back plane
(P10)
The main processors communicate with the IPMC using the Keyboard Controller Style (KCS)
interface. T wo KCS interfaces are available for the BIOS to communicate to the IPMC. BIOS uses
SMS interface for normal communication and SMM interface when executing code under systems
management mode (SMM). The base address of the LPC interface for SMS is 0xCA2 and 0xCA4
for SMM operation. Besides that, the BIOS is able to communicate with the IPMC for POST error
logging purposes, fault resilient purposes, and critical interrupts via the KCS interface.
The memory subsystem of the IPMC consists of a flash memory to hold the IPMC operation code,
firmware update code, system event log (SEL), and a sensor data record (SDR) repository. RAM is
used for data and occasionally as a storage area for code when flash programming is under
execution. The field replacement unit (FRU) inventory information is stored in the nonvolatile
memory on ADM1026. The flash memory can store up to 64 KBytes of SEL events and SDR
information, while the ADM1026 can store up to 512 bytes of FRU information. Having the SEL
and logging functions managed by the IPMC helps ensure that ‘post-mortem’ logging information
is available even if the system processor becomes disabled.
Intel NetStructure
Technical Product Specification
®
MPCBL0001 High Performance Single Board Computer 27
Hardware Management Overview
The IPMC provides six I2C bus connections. Two are used as the redundant IPMB bus connections
to the backplane while another one is used for communication with the ADM1026. The remaining
buses are unused. If an IPMB bus fault or IPMC failure occurs, IPMB isolators are used to switch
and isolate the backplane/system IPMB bus from the faulted SBC board. Where possible, the
IPMC activates the redundant IPMB bus to re-establish system management communication to
report the fault.
The onboard DC voltages are monitored by the ADM1026 device, manufactured by Analog
Devices. The IPMC queries the ADM1026 over a local system management I
ADM1026 includes voltage threshold settings that can be configured to generate an interrupt to the
IPMC if any of the thresholds are exceeded.
To increase the reliability of the MPCBL0001 SBC, a watchdog timer is implemented, whereby it
strobes an external watchdog timer at two-second intervals to ensure continuity of operation of the
board’s management subsystem. If the IPMC ceases to strobe the watchdog timer, the watchdog
timer isolates the IPMC from the IPMBs and resets the IPMC. The watchdog timer expires after six
seconds if strobes are not generated, and it resets the IPMC. Detailed information on the watchdog
timer configuration can be queried using standard IPMI v1.5 watchdog timer commands. The
watchdog timer does not reset the payload power.
3.1Sensor Data Record (SDR)
Sensor Data Records contain information about the t ype and number of sensors in the baseboard,
sensor threshold support, event generation capabilities, and the types of sensor readings handled by
system management firmware.
The MPCBL0001 management controller is set up as a satellite management controller (SMC). It
does support sensor devices, whose population is static by nature. SDRs can be queried using
Device SDR commands to the firmware. Refer to Section B, “List of Supported Commands (IPMI
v1.5 and PICMG 3.0)” on page 163 for the list of supported IPMI commands for SDRs. Hardware
sensors that have been implemented are listed below.
Table 2. Hardware Sensors (Sheet 1 of 3)
2
C bus. The
Sensor
Number
03hWatchdog Timer IPMC Watchdog
06hSystem Firmware
07hCPU Critical
08hMemory ErrorECC Multiple Bit
09hPower UnitPayload PowerIPMCPower OnSoft power control
NOTE: The PROCHOT signal is a discrete signal but it is treated as a threshold sensor so that it can have a
Sensor Type
Sensor Type of Temperature. IPMI does not have a discrete sensor type for temperatures. The
advantage of the PROCHOT sensor acting as a temperature sensor is that the CMM can recognize
events from this sensor as temperature events and adjust fan speed accordingly.
Volt age/Signals
Monitored
IPMB-0
assertion of the SMI
line
Monitored
via
LogicalPower On/
IPMCPower OnSMI Line asserted
Scanning
Enabled
under Power
State
Off
Off
Health LED
(Green to Red)
threshold
threshold
threshold
threshold
No change
No change
(Offset bit 01h asserted)
3.2System Event Log (SEL)
The SEL is the collection of events that are generated by the IPMC. Event logs are stored in nonvolatile memory. This resides on the board and allows better tracking of error conditions on the
baseboard when it is moved from chassis to chassis. Having the SEL and logging functions
managed by the IPMC helps ensure that post-mortem logging information is available should a
failure occur that disables the systems processor(s). In the MPCBL0001, flash memory for IPMI
firmware can store up to 3276 SEL entries. Management software running on the host processor is
responsible for ensuring that SEL storage has sufficient space for SEL logging. Events are
normally forwarded to shelf manager and logged to SEL on the board. If SEL storage on the board
is full, new events are forwarded to the Shelf Manager but are not logged in to SEL on the board.
A set of IPMI commands (see T able 97, “IPMI 1.5 Supported Commands” on page 163) allows the
SEL to be read and cleared and allows events to be added to the SEL. The IPMI commands used
for adding events to the SEL are Platform Event Message, Add SEL entry, and Partial Add Entry.
T able 3, “SEL Events Supported by the MPCBL0001 SBC” on page 31 lists supported SEL events.
Event Messages can be sent to the IPMC via the IPMB. This provides the mechanism for satellite
controllers to detect events and log them into the SEL.
30Intel NetStructure
®
MPCBL0001 High Performance Single Board Computer
Technical Product Specification
Hardware Management Overview
Table 3. SEL Events Supported by the MPCBL0001 SBC (Sheet 1 of 4)
Sensor
Type
Reserved00h-Reserved-
Temperature 01h-TemperatureThreshold exceeded for upper critical, upper non-
Power Unit09h00hPower Off/Power OnNormal power off indication. Offset 0 is just a status
Memory0Ch00hCorrectable ECCEvent data 3 = DIMM pair number
NOTE:
1. These sensor offsets do not generate events, but they are valid offsets when reading the sensor values.
Sensor
Type Code
Sensor-Specific
Offset (Event
Data 1, Bit 0-3)
01hThermal TripProcessor thermal trip has occurred.
04hFRB3/Processor Startup/
Initialization Failure
(CPU did not start)
05hConfiguration ErrorCPU 0 and CPU 1 are not present.
07hProcessor Presence
Detected
09hTerminator Presence
Detected
05hSoft Power Control
Failure (unit did not
respond to request to
turn on)
01hUncorrectable ECCEvent data 3 = DIMM pair number
EventRemarks
critical, lower critical and lower non-critical
thresholds. Refer to Table 4, “Sensor Thresholds for
IPMC Firmware 1.0” on page 35 for sensor
thresholds data.
lower critical and lower non-critical thresholds. Refer
to Table 4 for sensor thresholds data.
An FRB3 Timer (30 seconds) was implemented to
detect the failure of the CPUs from booting.
Event data 3 = Last Post 80 code byte
1
1
indicating that the payload power is off. It does not
generate an event when it is set. (For internal use).
The Power Unit sensor is used to detect when the
Payload power does not come up when the board is
told to power on.
When the board enters M4 state, the IPMC asserts a
Power Enable line to cause the Payload to power up.
The IPMC then waits for another line that indicates
that the power has come up successfully. If that line
does not assert within 2 seconds, then offset 05h is
asserted on the Power Unit sensor, which generates
an event to notify the Shelf Manager of the failure.
00 refers to J8/J9
01 refers to J10/J11
00 refers to J8/J9
01 refers to J10/J11
Intel NetStructure
®
MPCBL0001 High Performance Single Board Computer 31
Technical Product Specification
Hardware Management Overview
Table 3. SEL Events Supported by the MPCBL0001 SBC (Sheet 2 of 4)
Sensor
Type
System
Firmware
Progress
Event
Logging
Disabled
NOTE:
1. These sensor offsets do not generate events, but they are valid offsets when reading the sensor values.
Sensor
Type Code
0Fh00hBIOS checksum errorEvent data 2 = 99h
10h00hCorrectable Memory
Sensor-Specific
Offset (Event
Data 1, Bit 0-3)
EventRemarks
Event data 3 = 99h
Timer Count Read/Write
error
CMOS Battery errorEvent data 2 = FEh
CMOS Diagnosis status
error
CMOS Checksum errorEvent data 2 = FEh
CMOS Memory Size
error
RAM Read/Write test
error
CMOS Date/Time errorEvent data 2 = FEh
Clear CMOS jumperEvent data 2 = FEh
Clear Password JumperEvent data 2 = FEh
Manufacturing JumperEvent data 2 = FEh
Configuration error on
DIMM pair 0 (J8 & J9)
Configuration error on
DIMM pair 1(J10/J11)
No system memory is
physically installed or
fails to access any
DIMM's SPD data
BMC in update errorEvent data 2 = FEh
BMC Response Fail
error
Event Log Full errorEvent data 2 = FEh
Error Logging Disabled
Event data 2 = FEh
Event data 3 = 00h
Event data 3 = 01h
Event data 2 = FEh
Event data 3 = 02h
Event data 3 = 03h
Event data 2 = FEh
Event data 3 = 04h
Event data 2 = FEh
Event data 3 = 05h
Event data 3 = 06h
Event data 3 = 07h
Event data 3 = 08h
Event data 3 = 09h
Event data 2 = FEh
Event data 3 = 10h
Event data 2 = FEh
Event data 3 = 11h
Event data 2 = FEh
Event data 3 = 12h
Event data 3 = 0Ah
Event data 2 = FEh
Event data 3 = 0Bh
Event data 3 = 0Ch
Error Logging will be disabled after 10 events within
one hour.
32Intel NetStructure
®
MPCBL0001 High Performance Single Board Computer
Technical Product Specification
Hardware Management Overview
Table 3. SEL Events Supported by the MPCBL0001 SBC (Sheet 3 of 4)
Sensor
Type
Critical
Interrupt
Sensor
Type Code
13h04hPCI PERREvent data 2 = Bus No.
Sensor-Specific
Offset (Event
Data 1, Bit 0-3)
EventRemarks
Event data 3:
Byte [7:3] = Device No
Byte [2:0] = Func. No
05hPCI SERREvent data 2 = Bus No.
Event data 3:
Byte [7:3] = Device No
Byte [2:0] = Func. No
07hPCI Non-Fatal errorEvent data 2 = Bus No.
Event data 3:
Byte [7:3] = Device No
Byte [2:0] = Func. No
System
ACPI Power
state
22h00hS0/G0
06hS4/S5
0BhLegacy ON state
0ChLegacy OFF state
1
1
1
1
Board is running
Soft-off
Indicate ON for board that doesn’t support ACPI
Legacy soft-off
Watchdog23h00hTimer expired, status
only
01hHard ResetPOST/Boot monitor timed out
02hPower DownOS WDT shutdown after the monitor timeout
03hPower CycleOS WDT reset after the monitor timeout
08hTimer InterruptEvent data 2:
Byte [7:4] = Interrupt Type
0h = none
2h = NMI
Boot Error1Eh03hInvalid Boot SectorEvent will be logged if there’s an invalid boot sector
detected by the BIOS.
SMI Timeout E0h00hState De-Asserted
1
This is the normal situation when a board is able to
power up.
01hState AssertedThe SMI line has been constantly asserted for 10
seconds which indicates a severe hardware failure
around the CPU.
NOTE:
1. These sensor offsets do not generate events, but they are valid offsets when reading the sensor values.
Intel NetStructure
®
MPCBL0001 High Performance Single Board Computer 33
Technical Product Specification
Hardware Management Overview
Table 3. SEL Events Supported by the MPCBL0001 SBC (Sheet 4 of 4)
Sensor
Type
FRU Hot
Swap
IPMB Link
Sensor
NOTE:
1. These sensor offsets do not generate events, but they are valid offsets when reading the sensor values.
Sensor
Type Code
F0h00hM0 – FRU not installedRefer to PICMG 3.0 Specifications (Table 3-14)
F1h00hIPMB A & B disabledRefer to PICMG 3.0 Specifications (Table 3-46)
Sensor-Specific
Offset (Event
Data 1, Bit 0-3)
01hM1 – FRU inactive
02hM2 – FRU activation
request
03hM3 - FRU activation in
progress
04hM4 - FRU active
05hM5 - FRU deactivation
request
06hM6 - FRU deactivation in
progress
07hM7 - Communication lost
01hIPBM A enabled
IPMB B disabled
02hIPMB A disabled
IPMB B disabled
03hIPMB A & B enabled
EventRemarks
3.2.1Temperature and Voltage Sensors
Temperature and voltage readings are monitored by ADM1026. They are critical sensors that
ensure the MPCBL0001 is operating at its predefined threshold limits. The sensors are categorized
as follows:
• Lower Non-Critical
• Lower Critical
• Upper Non-Critical
• Upper Critical
If the lower critical or upper critical threshold is exceeded, it raises a major alarm. If the lower noncritical or upper non-critical threshold is exceeded, it raises a minor alarm.
Only critical thresholds which are exceeded turn the Health LED solid red. However, for any
events above, IPMC forwards the events to the shelf manager to log it into shelf manager ’s SEL.
34Intel NetStructure
®
MPCBL0001 High Performance Single Board Computer
Technical Product Specification
Table 4. Sensor Thresholds for IPMC Firmware 1.0
Hardware Management Overview
Sensor Name
+1.5 V1DhYes+1.5 VTBD1.431.451.551.57–
+2.5 V17hYes+2.5 VTBD2.32.362.6252.7–
+1.8 V16hYes+1.8 VTBD1.711.7461.8541.89–
VTT DDR
(+1.25 V)
+1.2 V14hYes+1.2 VTBD1.141.1761.2241.26–
+5 V19hYes+5 VTBD4.74.855.255.275–
-12 V1BhYes-12 VTBD-13.2-12.6-11.4-10.8–
+12 V1AhYes+12 VTBD10.811.412.613.2–
CPU Core
Voltage
+3.3 V18hYes+3.3 VTBD3.102 3.2013.4653.482–
+1.8 VSB12hYes+1.8 VTBD1.711.731.8361.89–
+3.3 VSB10hYes+3.3VTBD3.102 3.2013.4653.482–
+5 VSB11hYes+5 VTBD4.094.195.255.275–
VBAT13hYes+3 VTBD2.02.43.43.6–
Board
Temperature
CPU 0
Temperature
CPU 1
Temperature
NOTE: The following terms apply:
LNR: Lower Non-Recoverable
MPCBL0001 High Performance Single Board Computer 35
Technical Product Specification
Hardware Management Overview
Table 5. Sensor Thresholds for IPMC Firmware 1.2
Sensor NameDescription
+1.5V+1.5V1Dh1.51.43
+2.5V+2.5V17h2.492.29
+1.8V+1.8V16h1.791.71
VTT DDRDDR Voltage15h1.241.19
+1.2V+1.2V14h1.21.14
+5V+5V19h4.994.73
-12V-12V1Bh-12.11-15.06
+12V+12V1Ah12.17.56
CPU Core
Voltage
+3.3V+3.3V18h3.33.13
+1.8VSB+1.8V on
+3.3VSB+3.3V on
+5VSB+5V on
VBATBattery
Baseboard
Te mp
CPU 1 TempCPU 1 (Right)
CPU 2 TempCPU 1 (Left)
CPU Core
Voltage
standby rail
standby rail
Standby rail
voltage
Board
temperature
temperature
Temperature
Sensor
Number
1Ch1.311.24
12h1.791.71
10h3.33.13
11h54.09
13h3.551.99
30h30-5
37h405
38h405
Normal
Value
Lower
Critical
(1.45)
(2.32)
(1.73)
(1.16)
(1.16)
(4.78)
(-14.92)
(7.63)
(1.25)
(3.17)
(1.73)
(3.17)
(4.14)
(2.03)
(-2)
(8)
(8)
Thresholds
Lower
Noncritical
--1.57
2.35
(2.375)
--1.88
--1.31
--1.25
--5.23
-12.83
(-12.69)
11. 28
(11.313)
--1.37
--3.46
--1.88
--3.46
--5.24
3.31
(3.35)
5
(8)
10
(13)
10
(13)
Upper
Noncritical
2.63
(2.609)
-11.25
(-11.39)
12.85
(12.63)
---
60
(57)
76
(73)
76
(73)
Upper
Critical
(1.54)
2.69
(2.67)
(1.86)
(1.29)
(1.24)
(5.17)
-7.5
(-7.65)
15.06
(14.88)
(1.33)
(3.41)
(1.86)
(3.41)
(5.19)
70
(67)
81
(78)
81
(78)
Upper Nonrecoverable
-
-
-
-
-
-
-
-
-
-
-
-
80
(77)
127
(124)
127
(124)
NOTE: Values in parentheses are deassertion values.
36Intel NetStructure
®
MPCBL0001 High Performance Single Board Computer
Technical Product Specification
Hardware Management Overview
Table 6. Sensor Thresholds for IPMC Firmware 1.7 and Above
Sensor NameDescription
+1.5V+1.5V1Dh1.51.43
+2.5V+2.5V17h2.492.29
+1.8V+1.8V16h1.791.71
VTT DDRDDR Voltage15h1.241.19
+1.2V+1.2V14h1.21.14
+5V+5V19h4.994.73
-12V-12V1Bh-12.11-15.06
+12V+12V1Ah12.17.56
CPU Core
Voltage
+3.3V+3.3V18h3.33.13
+1.8VSB+1.8V on
+3.3VSB+3.3V on
+5VSB+5V on
VBATBattery
Baseboard
Te mp
CPU 0 TempCPU 0 - U35
CPU 1 TempCPU 1 - U36
CPU Core
Voltage
standby rail
standby rail
Standby rail
voltage
Board
temperature
Temperature
Temperature
Sensor
Number
1Ch1.311.24
12h1.791.71
10h3.33.13
11h54.09
13h3.551.99
30h30-5
37h405
38h405
Normal
Value
Lower
Critical
(1.45)
(2.32)
(1.73)
(1.16)
(1.16)
(4.78)
(-14.92)
(7.63)
(1.25)
(3.17)
(1.73)
(3.17)
(4.14)
(2.03)
(-2)
(8)
(8)
Thresholds
Lower
Noncritical
--1.57
2.35
(2.375)
--1.88
--1.31
--1.25
--5.23
-12.83
(-12.69)
11. 28
(11.313)
--1.37
--3.46
--1.88
--3.46
--5.24
3.31
(3.35)
5
(8)
10
(13)
10
(13)
Upper
Noncritical
2.63
(2.609)
-11.25
(-11.39)
12.85
(12.63)
---
60
(57)
76
(73)
76
(73)
Upper
Critical
(1.54)
2.69
(2.67)
(1.86)
(1.29)
(1.24)
(5.17)
-7.5
(-7.65)
15.06
(14.88)
(1.33)
(3.41)
(1.86)
(3.41)
(5.19)
70
(67)
81
(78)
81
(78)
Upper Non-
recoverable
-
-
-
-
-
-
-
-
-
-
-
-
80
(77)
127
(124)
127
(124)
NOTE: Values in parentheses are deassertion values.
Intel NetStructure
®
MPCBL0001 High Performance Single Board Computer 37
Technical Product Specification
Hardware Management Overview
Table 7. Sensor Thresholds for IPMC Firmware 1.14 and Above
Sensor NameDescription
+1.5V+1.5V1Dh1.51.43
+2.5V+2.5V17h2.492.29
+1.8V+1.8V16h1.791.71
VTT DDRDDR Voltage15h1.241.19
+1.2V+1.2V14h1.21.14
+5V+5V19h4.994.73
-12V-12V1Bh-12.11-7.46
+12V+12V1Ah12.17.56
CPU Core
Voltage
+3.3V+3.3V18h3.33.13
+1.8VSB+1.8V on
+3.3VSB+3.3V on
+5VSB+5V on
VBATBattery
Baseboard
Te mp
CPU 0 TempCPU 0 - U35
CPU 1 TempCPU 1 - U36
CPU Core
Voltage
standby rail
standby rail
Standby rail
voltage
Board
temperature
Temperature
Temperature
Sensor
Number
1Ch1.311.24
12h1.791.71
10h3.33.13
11h54.09
13h3.551.99
30h30-5
37h405
38h405
Normal
Value
Lower
Critical
(1.45)
(2.32)
(1.73)
(1.16)
(1.16)
(4.78)
(-7.61)
(7.63)
(1.25)
(3.17)
(1.73)
(3.17)
(4.14)
(2.03)
(-2)
(8)
(8)
Thresholds
Lower
Noncritical
--1.57
2.35
(2.375)
--1.88
--1.31
--1.25
--5.23
-11.21
(-11.35)
11. 28
(11.313)
--1.37
--3.46
--1.88
--3.46
--5.24
3.31
(3.35)
5
(8)
10
(13)
10
(13)
Upper
Noncritical
2.63
(2.609)
-12.79
(-12.65)
12.85
(12.63)
---
60
(57)
76
(73)
76
(73)
Upper
Critical
(1.54)
2.69
(2.67)
(1.86)
(1.29)
(1.24)
(5.17)
-14.95
(-14.81)
15.06
(14.88)
(1.33)
(3.41)
(1.86)
(3.41)
(5.19)
70
(67)
81
(78)
81
(78)
Upper Nonrecoverable
-
-
-
-
-
-
-
-
-
-
-
-
80
(77)
127
(124)
127
(124)
NOTE: Values in parentheses are deassertion values.
38Intel NetStructure
®
MPCBL0001 High Performance Single Board Computer
Technical Product Specification
Hardware Management Overview
3.2.2Processor Events
The processor asserts IERR as the result of an internal error. A thermal trip error indicates the
processor junction temperature has reached a level where permanent silicon damage may occur.
Upon THERMTRIP assertion, the IPMC powers down the boards.
3.2.3DIMM Memory Events
The MCH (E7501) instructs the ICH3 to report memory parity errors via SMI#. The SMI handler
extracts the error information (address) from the DRAM error registers in the MCH and logs it into
the SEL. The KCS interface performs error reporting to IPMC. BIOS sends a platform event
message with the appropriate data to the IPMC, which logs the event to SEL and forwards the
event to the Shelf Manager. Correctable memory errors generate an SMI and are logged into SEL.
Normally , a board with non-correctable errors is likely to hang as the multi-bit error may cause the
CPU to execute corrupted instructions. If the CPU executes corrupted instructions before executing
the code to log the event, then this event will not be logged in the SEL.
3.2.4System Firmware Progress (POST Error)
The BIOS is able to log both POST and critical events to the IPMC error log. (Refer to Table 83,
“BIOS Error Messages” on page 127.)
3.2.5Critical Interrupts
In general, the system BIOS is capable of generating requests on the KCS interface to
communicate with the IPMC for error logging, fault resilience, critical interrupts and reading/
writing inventory CPUs and RAM information to the IPMC. Two LPC interfaces are available for
the BIOS to communicate to the IPMC. The BIOS uses the SMS interface for normal
communication with the IPMC and the SMM interface when executing code under SMM mode.
PCI errors implemented in the MPCBL0001 are handled as follows:
1. The MCH(E7501) sends a parity error/system error (PERR/SERR) message over the hub
interface to the ICH3 notifying it that an error occurred.
2. The ICH3 generates an SMI# interrupt when it receives a PERR/SERR message.
3. The SMI handler checks the error status registers of CPU/MCH until it identifies the source
and type of the error.
4. The handler sends a message to the IPMC via the KCS interface, causing it to log the error in
the IPMC’s event log. IPMC then forwards the event to Shelf Manager to log it into Shelf
Manager SEL.
Intel NetStructure
Technical Product Specification
®
MPCBL0001 High Performance Single Board Computer 39
Hardware Management Overview
Table 8 shows the PCI mapping of the component subsystem of the baseboard.
Table 8. PCI Mapping for Hardware Component Subsystem
BusDeviceFunctionHardware Component Subsystem
000E7501 MCH Bridge
001MCH <-> ICH3
02082870P2 PCI-X Bridge (PMC and Gigabit Ethernet Controller)
Comments: FromTable 3 on page 31, event data 1, bit 3:0 is referring to PCI-PERR
• Event data 2 = 04.
Comments: From Table 3, event data 2, bit 7:0 is referring to Bus nu mber 4.
• Event data 3 = 08 = 00001000
Comments: FromTable 3, event data 3, bit 7:3 is equivalent to 1 whic h refers to Device nu mber 1.
Event data 3, bit 2:0 is equivalent to 0 which refers to Function number 0. From Table 8 above, the
PCI parity error was on the interface of the Gigabit Ethernet Controller (Port A).
40Intel NetStructure
®
MPCBL0001 High Performance Single Board Computer
Technical Product Specification
3.2.6System ACPI Power State
MPCBL0001 is targeted to support ACPI functionality, with support for the sleep states S0, S4 &
S5. On assertion of ICH3_SLP_S5# and ICH3_SLP_S3# GPIOs, IPMC sends out a hot-swap event
message to the shelf manager requesting deactivation. On successful reception of a deactivation
message from the shelf manager, the FRU enters M1 power state and remains in this state.
Under conditions where an ACPI enabled operating system is in S4/S5 sleep state, the chipset
could deassert ICH3_SLP_S5# and ICH3_SLP_S3# GPIOs requiring the IPMC to attempt
AdvancedTCA power state transition to M4 state (through M2, M3).
ACPI capabilities of an operating system are communicated by BIOS to the IPMC at initialization.
An OEM style IPMI command is sent by BIOS for this purpose. This command (SetACPIConfig ;
NetFn: 30h, command: 83h) is sent by BIOS every time an operating system is initialized. The
IPMC firmware defaults to no ACPI until this command is received with proper data in the request
to indicate the OS is either ACPI enabled or disabled. For obvious reasons, this command is only
executable over SMS channel.
3.2.7IPMB Link Sensor
The MPCBL0001 provides two IPMB links to increase communication reliability to the shelf
manager and other IPM devices on the IPMB bus. These IPMB links work together for increased
throughput where both busses are actively used for communication at any point. A request might
be received over IPMB Bus A, and the response is sent over IPMB Bus B. Any requests that time
out are retried on the redundant IPMB bus. In the event of any link state changes, the events are
written to the MPCBL0001 SEL. IPMC monitors the bus for any link failure and isolates itself
from the bus if it detects that it is causing errors on the bus. Events are sent to signify the failure of
a bus or, conversely, the recovery of a bus.
Hardware Management Overview
3.2.8FRU Hot Swap
The hot-swap event message conveys the current state of the FRU, the previous state, and a cause
of the state change as can be determined by the IPMC. Refer to PICMG 3.0 Specifications for
further details on the hot-swap state.
3.2.9CPU Failure Detection
A CPU failure during runtime or POST will have better error handling: a SEL event notification
will be generated if either one of the CPUs fails to power up, and the Health LED will turn red.
1. An FRB3 timer (30 seconds) was implemented to detect the failure of the CPUs to boot. This
also now implements offset 04h in the CPU 0 Status sensor . When asserted, it will generate an
event and set the Health LED to red.
2. The SMI line is now checked for a long (10 second) assertion that indicates a severe hardware
failure around the CPUs during runtime. As a result, a new discrete sensor has been added
(SMI Timeout) that will assert when the SMI line stays asserted too long.
Refer to Table 9 for the SEL events associated with FRB3 timer timeout and SMI Timeout
assertion.
Intel NetStructure
Technical Product Specification
®
MPCBL0001 High Performance Single Board Computer 41
Hardware Management Overview
Table 9. CPU Failure Behavior
CPU Failure DetectionCPU IdentificationBehavior
Operational PhaseCPU0CPU1Board Power Status
POSTNormalNormalBootable NoGreen
FailNormalStop BootingYesRed
NormalFailStop BootingYesRed
FailFailStop BootingYesRed
RuntimeNormalNormalKeep WorkingNoGreen
FailNormalHaltYesRed
NormalFailHaltYesRed
FailFailHaltYesRed
3.2.10Port 80h POST Codes
When there is an FRB3 failure, the event message sent from the CPU Status sensor with sensor
type code 07 provides the last Port 80 code byte written by the BIOS. This information is contained
in Data Byte 3 of the event message.
Example:
To decode Port 80 data from SEL event when a board is booted without memory, refer to the
following method.
SEL EVENT - ID:0DD8(Tue Jan 25 18:45:20 2005) Gen:8E Type:07 No:50 Dir:6F D1:64 D2:6F
D3:E1
CMM SEL
Event
Health LED
The values shown in bold above convey the following information:
• The sensor type is 07. This refers to the processor.
• Event data 1, bit 0-3 is 4. This refers to an FRB3/processor startup or initialization failure (the
CPU did not start).
• Event data 3 is E1. This refers to the last Port 80h POST codes before the board hangs.
Refer to the tables in Section 9.2 for descriptions of the Port 80h POST codes.
Note: At any time when a board hangs, you may also use an OEM IPMI command to query the Port 80
POST codes. For the command syntax, refer to Section 3.7.7 , “Get P ort80 Data” on page 50.
42Intel NetStructure
®
MPCBL0001 High Performance Single Board Computer
Technical Product Specification
Hardware Management Overview
3.3Field Replaceable Unit (FRU) Information
The FRU Information provides inventory data about the boards where the FRU Information Device
is located. The part number or version number can be read through software.
FRU information in the MPCBL0001 includes data describing the MPCBL0001 board as per
PICMG 3.0 Specification requirements. Additional multirecords will be added for the BIOS to
write CPU information, BIOS version number, and PMC information to FRU data correctly. This
information is retrieved by shelf manager (ShMC), enabling reporting of board-specific
information through an out-of-band mechanism.
Following are the definitions for the multirecord implemented by the firmware as part of FRU data.
Table 10. FRU Multirecord Data for CPU/RAM/PMC/BIOS Version Information
VariableSize (byte)DataType
Manufacturer ID
(Intel IANA number)
Record Version11Binary
Type/Length11Binary
CPU No.s1xBinary
Type/Length12Binary
RAM Info2X (in units of 1 MByte)Binary
Type/Length1(5 * XXX) + 1Binary
No. of PMCs1XXXBinary
PMC Info5*XXXPMC_DataBinary
Type/Length10xFFBinary
BIOS Version63 (max)yyyyyyyyASC-II
End of fields10xC1Binary
PMC Installed?10 (PMC is not installed)
+
30x000157
(LSB first, MSB next)
1 (PMC is installed)
Binary
Binary
Table 11. PMC Data
VariableSizeDataType
Device ID2XXBinary
Vendor ID2XXBinary
Intel NetStructure
Technical Product Specification
®
MPCBL0001 High Performance Single Board Computer 43
Hardware Management Overview
3.4E-Keying
E-Keying has been defined in the PICMG 3.0 Specification to prevent board damage, prevent
misoperation, and verify fabric compatibility. The FRU data contains the board point-to-point
connectivity record as described in Section 3.7.2.3 of the PICMG 3.0 Specification.
Upon management power-on, the firmware sets the Fibre Channel ports to front panel by default.
When the board enters M3 power state, the shelf manager reads in the board point-to-poin t
connectivity record from FRU and determines whether the board can enable the Fibre Channel
ports to the back plane. Set/Get Port State IPMI commands defined by the PICMG 3.0
Specification are used for either granting or rejecting the E-keys.
If user Fibre Channel selection is to the front, the firmware maintains the Fibre Channel ports to the
front panel regardless of the shelf manager’s granting or rejecting of E-keys for the board.
Table 12 on page 44, describes the:
• Connections to base and fabric interfaces on the MPCBL0001 board for E-keying purposes.
• Link descriptor list for the two Gigabit Ethernet channels connected to the base interface and
the two Fibre Channels on the fabric interface.
Table 12. Link Descriptors for E-Keying
NoLink
Descriptor
1Ethernet
Port 1
2Ethernet
Port 2
3FC Port 10001000000010 1000010000010x00202C41
4FC Port 20001000000010 1000010000100x00202C42
NOTE: Fibre Channel E-keying is only applicable to MPCBL0001FXX products.
Link
Grouping
ID
[31:24][23:20][19:12]{11:8}{7:6}[5:0}
0000000000001 0001000000010x00001101
0000000000001 0001000000100x00001102
Link T ype
Extension
3.5IPMC Firmware Code
IPMC firmware code is organized into boot code and operational code, both of which are stored in
a flash module. Upon an IPMC reset, the IPMC executes the boot code and performs the following:
1. Self test to verify the status of its hardware and memory.
2. Sets up the internal real-time operating system (RTOS).
3. Performs a checksum of the operational code.
Upon successful verification of the operational code checksum, the firmware will jump to the
operational code.
Link TypeLink DesignatorLink Desc
Port 0 3 Flags
InterfaceChannel
Number
Value
44Intel NetStructure
®
MPCBL0001 High Performance Single Board Computer
Technical Product Specification
When the firmware is commanded to enter firmware (FW) update mode, the operational code uses
a special branch, Software Interrupt, to jump to the FW update code in the boot block. Once in FW
update mode, the update code is copied into RAM, then the FW jumps to the code in RAM to
execute. The FW update code cannot execute out of flash while the flash is being updated.
Figure 4. IPMC Firmware Code Process
IPMC Boot Block
Hardware Management Overview
!
!
"#
Main IPMC Code
RAM
"#
"
No
No
!
$"
3.6IPMC Firmware Upgrade Procedure
MPCBL0001 firmware is upgraded using either of two methods, the KCS interface or the IPMB
(RMCP) interface.
3.6.1IPMC Firmware Upgrade Using KCS Interface
The KCS interface is the communication mechanism between the host processor on the
MPCBL0001 and the IPMC controller. A firmware update utility is available. It takes a hex file to
be updated as input from the command line. It can also verify that updates are completed
successfully by reading back data written to the flash memory. Typically, it takes the utility around
two minutes to complete the update over the KCS interface. After the firmware update is
Intel NetStructure
Technical Product Specification
®
MPCBL0001 High Performance Single Board Computer 45
Hardware Management Overview
completed, the controller goes through a reset and boots up with the new firmware. The host
processor is not reset when going through a firmware update, so the operating system and
applications running on the host processor are not interrupted.
Below is a step-by-step procedure on how to update the firmware:
1. Copy the FW upgrade utility and FW upgrade (Hex) file to a DOS bootable floppy disk.
2. Boot MPCBL0001 from a USB floppy disk (connected to the USB port) to a DOS prompt.
3. Copy the automatically generated (C: drive) Upgrade utility and hex file to RAM disk.
4. Issue the command “FWPIAUPD filename.hex” (or whatever the actual hex file name is).
5. MPCBL0001 is now in FW upgrade mode. Select 'U' for updating firmware or 'V' to verify the
hex image on the board with the image being used by the update utility.
6. The utility shows the versions on the board and on the hex image being updated.
7. Select 'Y' to update; follow the prompts and answer appropriately.
The upgrade ends with a message of successful termination.
3.6.2IPMC Firmware Upgrade via the IPMB Interface (RMCP)
Figure 5. Upgrade via Remote Management Node
Remote
Management
Node
(RMCP Client)
LAN
IPMI Specification v1.5 defines Remote Management Control Protocol (RMCP). Version 1.5 adds
features for layering commands through virtual networks like Ethernet.
The IPMC firmware that needs to be upgraded is loaded to client utility software on the RMCP
client. The RMCP client uses the RMCP protocol carrying embedded IPMI messages to send to the
RMCP Server running in the CMM. The RMCP server decodes the RMCP package and forwards
the IPMI messages to the SBC.
Shelf
Management
(RMCP Server)
Intel® NetStructure™
MPCBL0001 High-
Performance SBC
IPMC
Intel NetStructure
MPCBL0001 High-
Performance SBC
IPMC
Intel NetStructure
MPCBL0001 High-
Performance SBC
IPMC
B2643-01
46Intel NetStructure
®
MPCBL0001 High Performance Single Board Computer
Technical Product Specification
3.6.2.1Updating MPCBL0001 Firmware
To update the MPCBL0001 firmware for the Intel NetStructure® MPCBL0001 SBC, execute the
following commands.
1. Copy the update utility (fwpiaupd) and the firmware image file into the same directory in the
RedHat Linux host. Note: If using ftp for file transfer, use binary mode to transfer files. The
firmware image file or the utility file may get corrupted if binary mode is not used.
Note: The dot before / character in the command is required.
• IPAddress is the IP address of the CMM.
• At default, user is root. Password is cmmrootpass.
• slot# is slot 1 to 14 in the MPCHC0001 chassis.
• filename.hex is the firmware file.
Hardware Management Overview
Note: This "fwpiaupd" utility can only be supported on the Red Hat* Linux* 8.0 Platform. The Intel
NetStructure
®
MPCMM0001 Chassis Management Module is needed for this remote update.
The utility and upgraded firmware are part of the IPMC Firmware release package. It can be
downloaded from the Intel web site at http://www.intel.com/design/network/products/cbp/atca/
index.htm.
3.7OEM IPMI Commands
This section documents the OEM style IPMI commands implemented and supported on the
MPCBL0001.
3.7.1Reset BIOS Flash Type
This command resets the processor and changes the BIOS bank select signal so that CPU boots off
redundant BIOS bank.
Table 13. Reset BIOS Flash Type
76543210
NetFn/LUNNetFn = 3Ah (OEM Request)RsLUN
CommandCmd = 01h
Byte 1BIOS checksum success/failure indication
00h – Checksum success
01h – Checksum failure
Byte 1Completion code
Intel NetStructure
Technical Product Specification
®
MPCBL0001 High Performance Single Board Computer 47
Hardware Management Overview
3.7.2Set Fibre Channel Port Selection
This command sets the Fibre Channel port routing as specified in the request data bytes. The
command is available over KCS and IPMB interface.
This command returns the current Fibre Channel port routing selection as set in the hardware. The
command is available over KCS and IPMB interface SetFiberChannelPortSelection.
MPCBL0001 High Performance Single Board Computer 49
Technical Product Specification
Hardware Management Overview
3.7.6Get Control State
This command sets the state of a control pin. This command overrides the AUTO-state of the
control pin. Refer to Table 20 on page 50 for control number in for mation.
This command returns the last byte value written by the BIOS to Port 80 since the last System
Reset. If no data has been written to the port since System Reset, the Completion Code returned is
CBh.
Table 19. Get Port80 Data
76543210
NetFn/LUNNetFn = 30h (OEM Request)RsLUN
CommandCmd = 2Dh
Byte 1— (BLANK)
Byte 1Completion code
Byte 2Last Port 80 code value (in HEX)
3.8Controls Identifier Table
Table 20 below lists the control identifiers that can be used with Set/Get Control State IPMI
commands to query or set information on certain controls in the firmware.
Table 20. Controls Identifier Table
Control DescriptionControl Number
FWH Hub (for BIOS bank information) 00
FWH 0 Write Protect1
FWH 1 Write Protect2
FWH 0 Top Block Lock3
FWH 1 Top Block Lock44
50Intel NetStructure
®
MPCBL0001 High Performance Single Board Computer
Technical Product Specification
3.9Hot-Swap Process
The MPCBL0001 SBC has the ability to be hot-swapped in and out of a chassis. The onboard
IPMC manages the SBC’s power-up and power-down transitions. The list below, along with
Figure 6, illustrates this process.
1. Ejector latch is opened. HOT_SWAP_PB# assertion. IPMC firmware detects the assertion of
this signal.
2. IPMC sends "Deactivation Request" message to CMM. M state moves from M4-> M5.
3. Board moves from M5 -> M6 if the CMM grants the request.
4. The IPMC's ACPI timer (3 minutes) starts if an ACPI-enable OS is loaded. Otherwise, it goes
to Step 7 below. The IPMC asserts 20 ms pulse on SMC_PWRBTN#.
5. The Power Button Status register (PWRBTN_STS) is set. It then asserts SCI/SMI# to the OS.
If ACPI OS is enabled, SCI interrupt handler on the OS is called. Interrupt handler clears
PWRBTN_STS bit. OS starts to perform a graceful shutdown.
6. ICH3 detects "LOW" on the ICH3_PWRBTN#. Asserts ICH3_SLP_S3# and ICH3_SLP_S5#
to IPMC. Upon detection of ICH3_SLP_S5# and ICH3_SLP_S3#, board transitions to Step 7
below. If ICH3 doesn't assert the signals, the board will transition to Step 7 below upon the
ACPI timer expiration.
7. The firmware deasserts payload power and sets the IPMI locked bit before it transitions from
M6 to M1 state.
Hardware Management Overview
Note: If the upper-level software moves the IPMC to M6, the same procedure is followed, starting with
Step 4.
Figure 6. Hot-Swap Process
ICH3
ACPI-OS
4
5
6
IPMC
1
CMM
2
3
7
Intel NetStructure
Technical Product Specification
®
MPCBL0001 High Performance Single Board Computer 51
Hardware Management Overview
3.9.1Hot-Swap LED (DS10)
The MPCBL0001 SBC supports one blue Hot Swap LED, mounted on the front panel. See
Figure 14, “MPCBL0001NXX SBC Front Panel” on page 71 for its location. This LED indicates
when it is safe to remove the SBC from the chassis. The on-board IPMC drives this LED to
indicate the hot-swap state. Refer to Table 21, “Hot-Swap LED (DS11)” on page 52.
When the lower ejector handle is disengaged from the faceplate, the hot swap switch embedded in
the PCB will assert a "HOT_SWAP_PB#" signal to the IPMC, and the IPMC will move from the
M4 state to the M5 state. At the M5 state, the IPMC will ask the CMM (or Shelf Manager) for
permission to move to the M6 state. The Hot Swap LED will indicate this state by blinking on for
about 100 milliseconds, followed by 900 milliseconds in the off state. This will occur as long as the
SBC remains in the M5 state. Once permission is received from the CMM or higher-level software,
the SBC will move to the M6 state.
The CMM or higher level software can reject the request to move to the M6 state. If this occurs, the
Hot Swap LED returns to a solid off condition, indicating that the SBC has returned to M4 state.
If the SBC reaches the M6 state, either through an extraction request through the lower ejector
handle or a direct command from higher-level software, and an ACPI-enabled OS is loaded on the
SBC, the IPMC communicates to the OS that the module must discontinue operation in preparation
for removal. The Hot Swap LED continues to flash during this preparation time, just like it does at
the M5 state. When main board power is successfully removed from the SBC, the Hot Swap LED
remains lit, indicating it is safe to remove the SBC from the chassis.
Warning:Removing the SBC prematurely can lead to device corruption or failure.
Table 21. Hot-Swap LED (DS11)
LED StatusMeaning
OffNormal status
Blinking BluePreparing for removal/insertion: Long blink indicates activation is in
Solid BlueReady for hot swap
progress, short blink when deactivation is in progress.
3.9.2Ejector Mechanism
In addition to captive retaining screws, the MPCBL0001 SBC has two ejector mechanisms to
provide a positive cam action; This ensures the blade is properly seated. The bottom ejector handle
also has a switch that is connected to the IPMC to determine if the board has been properly
inserted.
52Intel NetStructure
®
MPCBL0001 High Performance Single Board Computer
Technical Product Specification
3.10Interrupts and Error Reporting
3.10.1Device Interrupts
The Low Voltage Intel® Xeon™ processor and E7501 chipset (MCH, ICH3, P64H2) utilize a
mechanism for delivering interrupts that is slightly different from, though fully compatible with,
previous IA-32 system platforms. The change affects only the delivery mechanism and no changes
are required to existing software.
This new delivery mechanism transfers the equivalent APIC messages across the system bus
structure rather than using a sideband channel as in the case of the APIC serial bus. There is no
longer an APIC bus connection to the processor. This new mechanism improves the interrupt
message transfer speed to the processors, thus reducing latency. It also simplifies the flushing of
buffers that is required when data is buffered between the I/O subsystem and memory. Since
interrupt messages are no longer communicated across a sideband channel, these transfers are now
visible to the chipset. The interrupt message transactions themselves can now initiate buffer
flushing to ensure all data within the I/O and memory subsystems is coherent.
As before, the LINT[1:0] connections to the processors remain for compatibility with the old PC
industry standard, legacy interrupt architecture (8259 controllers). In addition, the P64H2 PCI
bridge devices include an interrupt output (BTINTR#), which can be routed into the legacy
interrupt controller to facilitate booting from devices residing on the far side of such PCI bridge
devices. Once the boot process is complete and the APIC interrupt system is enabled, devices no
longer need to share interrupts; This improves interrupt system performance.
Hardware Management Overview
The BIOS initializes and enables both the 8259 and APIC but masks all APIC interrupts in the
redirection table. This is so the SBC operates in legacy interrupt mode. The BIOS does not operate
in APIC mode at any time. An APIC-aware OS disables the 8259 and unmasks the APIC interrupts
to switch to APIC mode.
Table 22 displays the interrupt connecti ons pro vid ed by the MPCBL0001 SBC. Actual interrupt
vector assignments and routing to legacy interrupts as necessary is under BIOS and/or OS control.
Table 22. Interrupt Assignments (Sheet 1 of 2)
Legacy InterruptIRQ assigned
Master 8259
Internal timer0 output0
Slave 8259 INTR output2
Serial Port A3
Slave 8259
Internal RTC0 (8)
Primary IDE6 (14)
PCI Device InterruptIRQ assigned
HI-A ICH3
Super I/OSERIRQ
USB 1.1 controller #1PIRQA#
IPMC_SYSIRQ#PIRQB#
Intel NetStructure
Technical Product Specification
®
MPCBL0001 High Performance Single Board Computer 53
Hardware Management Overview
Table 22. Interrupt Assignments (Sheet 2 of 2)
Legacy InterruptIRQ assigned
HI-B P64H2 BTINTR# PIRQC#
HI-C P64H2 BTINTR# PIRQD#
HI-B P64H2
Fibre Channel INTA#PB_IRQ0
Fibre Channel INTB#PB_IRQ1
HI-C P64H2
PMC INTA#PA_IRQ0
PMC INTB#PA_IRQ1
PMC INTC#PA_IRQ2
PMC INTD#PA_IRQ3
Ethernet #1 INTA#PB_IRQ0
Ethernet #2 INTA#PB_IRQ1
Figure 7. Interrupt Signals
Low Voltage
®
Intel
Xeon
Processor
MCH
8-Bit
HL_A
ICH3
10x
APIC
LPC
Super I/O
LINT0 / INTR
LINT0 / NMI
Low Voltage
®
Intel
Xeon
®
Processor
®
FSB
Redirection Logic
16-Bit
HL_B
P64H2
10x
APIC
PCI-X
Fibre
Channel
P64H2
10x
APIC
GbE
(x2)
16-Bit
HL_C
PCI-X
PMC
Slot
PCI
64/66
54Intel NetStructure
B0754-01
®
MPCBL0001 High Performance Single Board Computer
Technical Product Specification
3.10.2Error Reporting
The MCH handles error reporting from the memory subsystem. Errors consist of correctable and
uncorrectable bit errors. The ECC algorithms used are capable of correcting any number of bit
errors contained within a 4-bit nibble. In addition, any number of bit errors contained within two 4bit nibbles is detected. The MCH communicates these errors to the ICH3 via special cycles over the
hub link interface. These special cycles indicate to the ICH3 that an MCH-detected error has
occurred. The MCH special cycle communicates the type of event that should be generated by the
ICH3 when an error is detected. Selection for the generation of an SERR, SMI, or SCI event is
provided. Status for these reported errors is then found in the MCH DRAM_FERR (first error) and
DRAM_NERR (next error) status registers. Refer to the MCH data sheet for more information (see
Appendix A, “Reference Documents”).
Correctable memory errors generate an SMI and are logged via IPMI as a SEL. Non-correctable
errors first generate an SMI (which generates a SEL) and then an NMI.
Each P64H2 device reports the PCI errors that occur on the buses to which it is attached. These
consist of the PCI error assertions of the PERR# or SERR# signals. The errors are reported by
sending the DO_SERR special cycle to the MCH on the Hub Interface. The MCH forwards the
error to the ICH3, which generates the appropriate error condition to the processor(s) such as NMI,
SMI, or SCI.
PCI address parity errors are considered catastrophic and may abort further data transfers by the
P64H2 if that is the programmed response. Parity/ECC is checked on both the Hub Interface and
PCI bus transactions. PCI data parity errors are considered less severe and allow transactions to
continue. Data parity errors cause the Detected Parity Error” status to be logged and, if enabled, the
DO_SERR special cycle is transmitted. In a transaction where a data error occurs, the data being
forwarded to the next bus is “poisoned” to ensure the error follows the data to its destination.
Poisoned data has bad parity or multi-bit ECC errors introduced before being forwarded to the next
bus.
Hardware Management Overview
PCI assertions of the SERR# signal also result in the DO_SERR special cycle being generated on
the hub interface when enabled. Other potential causes for a DO_SERR special cycle include:
• Parity errors on the target bus during a write.
• A master timeout on a delayed transaction.
• The occurrence of a PCI master abort cycle.
Refer to the P64H2 Data Sheet, section 4.9, for more information on error handling. For details on
obtaining this document, see Appendix A, “Reference Documents.”
The ICH3 device has the ability to report PCI and hub link errors directly to the processors. When
a PERR# or SERR# occurs on the ICH3 local PCI bus, the ICH3 can be programmed to generate
NMI or SMI. The ICH3 also fields messages from the MCH and its attached hub devices to
indicate errors to the processors on their behalf. The messages may request SMI#, SCI, NMI, or
SERR3 to be asserted. Software must check the MCH and attached hub devices to determine the
exact cause of the error. Refer to the ICH3 Data Sheet for more information on error handling and
generation. For details on obtaining this document, see Appendix A, “Reference Documents.”
Intel NetStructure
Technical Product Specification
®
MPCBL0001 High Performance Single Board Computer 55
Hardware Management Overview
3.11ACPI
ACPI gives the operating system direct control over the power management and Plug and Play
functions of a computer. The use of ACPI with theMPCBL0001 SBC requires an operating system
that provides ACPI support. ACPI features include:
• Plug and Play (including bus and device enumeration) and APM support (normally contained
in the BIOS).
• Power management control of individual devices, add-in boards (some PMC cards may
require an ACPI-aware driver), and hard-disk drives.
• A soft-off feature that enables the operating system to power off the computer.
• Support for an IPMC firmware command switch.
3.11.1System States and Power States
Under ACPI, the operating system directs all system and device power state transitions. The
operating system puts devices in and out of low-power states based on user preferences and
knowledge of how devices are being used by applications. Devices that are not being used can be
turned off. The operating system uses information from applications and user settings to put the
system as a whole into a low-power state.
Table 23, “Power States and Targeted System Power” on page 56 lists the power states and the
associated system power targets supported by the MPCBL0001 SBC. See the ACPI Specification
for a complete description of the various system and power states.
3.12Reset Types
Table 23. Power States and Targeted System Power
Global StatesSleeping States
G0 – working stateS0 – workingC0 – workingD0 – working state.
G1 – sleeping stateS4 – Suspend to disk.
G2/S5S5 – Soft off. Context
G3 – mechanical off
AC power is disconnected
from the computer.
The watchdog timer on the IPMC can be configured and used through standard IPMI v1.5
watchdog timer commands. Refer to Section 3.13.1, “WDT #1” on page 62 for detailed
implementation.
Context saved to disk.
not saved. Cold boot is
required.
No power to the system. No powerD3 – no power for wake up
Processor
States
No powerD3 – no power except for wake
No powerD3 – no power except for wake
Device States
up logic.
up logic.
logic, except when provided by
battery or external source.
56Intel NetStructure
®
MPCBL0001 High Performance Single Board Computer
Technical Product Specification
3.12.1Reset Logic
The following topics describe the two types of reset requests and the boot relationships among
them. The two types of reset requests available on the MPCBL0001 are:
• Hard reset request (always results in a cold boot)
• Soft reset request (can result in either a warm or cold boot)
A hard reset request occurs whenever the processor Reset line is asserted and then deasserted. A
soft reset occurs whenever an assertion occurs on the processor Init line. Whenever a soft reset
request occurs, the BIOS checks two memory locations to determine whether to initiate a warm
boot while leaving main memory intact or a cold boot that clears memory.
Whenever the BIOS detects that the reset is either a hard reset or a cold boot, it specifically clears
the memory location 40h:72h so it does not contain a 1234h. Under warm boot conditions, this
memory location contains a 1234h (the developer’s application writes this value in this location
[using /dev/mem] when it is started). If a hard reset occurs (as defined in the hard reset topic
below), it is certain that the 40h:72h location contains a non-1234h value.
3.12.2Hard Reset Request
Hardware Management Overview
A Hard Reset, or CPU Reset, is defined as the assertion of the processor reset signal (see Table 24,
“Reset Request” on page 58). This initializes the processor state and registers, disables internal
caches, and causes the processor to unconditionally begin execution from the reset vector. A hard
reset is initiated by the following events:
1. A power up of the SBC. The SMC enables the onboard power supplies.
2. The SMC negates the ICH3_PWROK signal (see Note below).
3. A “reset” command from the Port CF9h I/O register (refer to the “Intel
Controller Hub 3 (ICH3-S) Datasheet” for information about this register).
4. Watchdog timer (WDT #1) expires and is configured to initiate a hard reset. See “Watchdog
Timers (WDTs)” on page 62 for more information.
5. Watchdog timer (WDT #3) expires after failure to perform the first instruction fetch.
6. A command (cmmset -l bladex -d powerstate -v reset) is issued from MPCMM0001.
Note: The IPMC can negate the dedicated signal ICH3_PWROK to initiate a processor reset.
ICH3_PWROK indicates whether power is OK. If the IPMC deasserts ICH3_PWROK, the
hardware asserts the processor reset lines.
3.12.3Soft Reset Request
The assertion of the processor’s INIT signal causes a soft reset or “CPU INIT” (see Table 24,
“Reset Request” on page 58). The ICH3 is normally responsible for driving th e INIT signal. A
CPU INIT event causes the processor(s) to fetch the reset vector at the next instruction boundary.
The majority of the processor and all of the cache states are unaffected by an INIT event.
®
82801CA I/O
After the INIT event, hardware may be reset (or not reset) under BIOS control. PCI buses are reset
using their respective bridge control registers. This signal is then level translated to the processor
compatible signal level. INIT may be caused by the following events:
Intel NetStructure
Technical Product Specification
®
MPCBL0001 High Performance Single Board Computer 57
Hardware Management Overview
1. The reset button is pressed (see Note below). See Sect ion 14, “MPCBL0001NXX SBC Front
Panel” on page 71 for its location.
2. A processor shutdown special cycle occurred.
3. An INIT command from Port 92h I/O register (refer to the Intel3 (ICH3-S) Datasheet for information about this register).
4. An INIT command from Port CF9h I/O register.
5. A keyboard reset command (ICH3 RCIN# signal asserted).
6. The IPMC may also directly assert the INIT signal; WDT #1 expires and is configured for a
soft reset.
7. Processor BIST is enabled and a hard reset is initiated from the Port CF9h register. This asserts
the INIT signal but is not classified as a soft reset since CPU reset is also asserted.
8. OS reboot commands (eg: "shutdown -r now" or "reboot" in Linux).
9. A processor INIT may also be initiated through an APIC “init” message. This message may
target a specific processor or all processors. This “init” is an internally generated event (No
INIT signal is asserted) so the IPMC is unable to detect this occurrence.
Note: The reset button (RESET_PB#) is an input to the IPMC. There are also IPMI commands to reset
the board and change power states through the software. However, the reset button is a last resort
because the user must be physically present at the chassis to reset the board.
®
82801CA I/O Controller Hub
After a Soft Reset/CPU Init, the BIOS code executes and determines if the reset is a warm boot or
a cold boot. A warm boot restarts the system and keeps memory above the 8 MByte boundary
intact. During a warm boot the MCH is not reset, allowing DRAM refresh to continue during and
over the soft reset event. A cold boot sets the state of all peripherals to the same state they would be
in if a hard reset were triggered.
Table 24. Reset Request
Reset RequestSignal ActivatedType
HardResetFull reboot
SoftInitPartial reboot
3.12.4Warm Boot
A warm boot occurs when the processor is booting after a soft reset request. To qualify as a warm
boot, the reset counter located at 40h:D0h must be non-zero (by default, the reset counter and reset
flag are initialized to 10 and 1234h by BIOS after a cold boot.) Execution starts at the reset vector.
The BIOS initializes and configures all devices except for memory. Memory contents remain intact
except for the first 8 MBytes. The BIOS uses the first 8 MBytes during POST, but does not modify
the reset flag or the reset counter. MCH is not reset, allowing DRAM refresh to continue during the
warm boot.
Note: On every warm boot, BIOS automatically decrements the reset counter by one. When the reset
counter reaches zero and the soft reset is initiated, a cold boot occurs instead of warm boot.
58Intel NetStructure
®
MPCBL0001 High Performance Single Board Computer
Technical Product Specification
3.12.5Cold Boot
Any soft reset that does not meet the configuration described in the preceding W arm Boot section is
classified as a cold boot. Execution starts at the reset vector, and BIOS initializes and configures all
devices, including memory subsystem, as if a hard reset had occurred. See Table 25, “Reset
Actions” on page 59.
During a cold boot the BIOS initializes the warm reset counter to 0x0A and clears the reset flag to
1234h. Software can then read the reset flag to determine the type of reset.
Cold bootFull restartFunctionally equivalent to a hard reset.
3.12.6Power Good
When the MPCBL0001 SBC is inserted into the chassis, the hardware management circuitry is
“hot plugged.” The hardware management voltage is immediately applied, and the on-board IPMC
is reset. After the hardware management reset, the operation of the IPMC and full power-up of the
SBC are under firmware control.
Hardware Management Overview
Upon command to power on the module, the IPMC asserts the “power enable” signal to the onboard DC/DC converters. Full power-up of the SBC is sequenced by hardware to ensure devicespecific power requirements are followed. Sequencing of specific voltages is required to ensure
that devices using multiple voltages are not damaged or stressed.
Figure 8. Power Good Map
VRM
Controller
Intelligent Platform
Management
Controller
(IPMC)
Power
Goods
PLD
VRM_PWRGOOD
ICH3_PWROK
ICH3
E7501
MCH
82546
Dual
GbEnet
H_PWRGD
ICH3_PCIRST#
Low Voltage
®
Xeon™
Intel
Processor
Low Voltage
®
Xeon™
Intel
Processor
ICH3_PCIRST#
(global reset)
Intel NetStructure
®
MPCBL0001 High Performance Single Board Computer 59
Technical Product Specification
B0895-02
Hardware Management Overview
As the many voltages power up, each regulator produces a “power good” signal. All of these power
good signals are logically OR’d (with the exception of the VRM power good) to produce the
ICH3_PWROK signal input to the ICH3 as shown in Figure 8, Power Good Map. When this signal
is active, it indicates all on-board power is good.
Next, the VRM power good is gated with the ICH3_PWROK signal in the ICH3 to produce the
processor’s power good signal input.
As soon as the ICH3 device is powered, its PCI reset output is asserted. This reset output remains
asserted until all power good signals are present (indicated by the ICH3_PWROK signal), the
processor VRM power good signal is asserted, and device voltage/clock stabilization times have
been satisfied.
Device resets are then released, and processor BIOS execution and boot begins. The PCI reset
output of the ICH3 is the source of all other power-up reset signals as shown in Figure 9, “Reset
Chain” on page 61
The IPMC is also capable of initiating this power-up or global reset by negating the
ICH3_PWROK signal. Additionally, devices on specific PCI buses may be independently reset by
software through their associated bridge devices.
When commanded to do so, the IPMC releases device and processor resets, and processor BIOS
execution and boot begins.
60Intel NetStructure
®
MPCBL0001 High Performance Single Board Computer
Technical Product Specification
Figure 9. Reset Chain
Hardware Management Overview
ICH3
ICH3_PCIRST#
FWH 0
FWH 1
E7501
H_CPURST
MCH
DIMMs
PCIRST A
HI-B
P64H2
PCIRST B
PCIRST A
HI-C
P64H2
PCIRST B
Super I/O
IPMC
(Intelligent Platform
Management
Controller)
Low Voltage
®
Xeon™
Intel
Processor
Low Voltage
®
Xeon™
Intel
Processor
<unused>
Fiber Channel
Controller
PMC card
82546
Dual
GbEnet
Intel NetStructure
®
MPCBL0001 High Performance Single Board Computer 61
Technical Product Specification
B0896-02
Hardware Management Overview
3.13Watchdog Timers (WDTs)
Figure 10, “Watchdog Timers” on page 62 shows the relationship between the three watchdog
timers (WDTs) on the MPCBL0001 SBC.
Figure 10. Watchdog Timers
ICH3
(South Bridge)
WDT #3
3.13.1WDT #1
The first WDT (WDT #1) is a hardware timer in the IPMC. WDT #1 is IPMI compliant; its
interaction with the host processor BIOS or system software is accomplished through IPMI
commands over the Keyboard Controller Style (KCS) interface to the IPMC. The host processor
uses the Set Wat chdog Timer message to configure WDT #1, then the Reset Watchdog Timer
message to strobe the timer.
Strobe
IPMB-A
Isolation Logic
Host
Processor(s)
IPMC
WDT #1
Strobe
Strobe
PLD
WDT #2
IPMB-B
Isolation Logic
B1368-02
WDT #1 can be set to any value between 100 ms and 6,553,600 ms in 100 ms intervals. Another
configuration parameter is an indicator of which software is controlling WDT #1. This has five
state settings:
1. BIOS FRB2: Used during fault-resilient booting to detect issues in the BIOS.
2. BIOS/POST: Used while the BIOS is running through its POST operations.
3. OS Load: Set by the BIOS just before an OS load, then reset by the OS (the OS must be
enabled to do so) when it finishes booting.
4. SMS/OS: Used by the system management software or the OS.
5. OEM: Used by any OEM software.
62Intel NetStructure
®
MPCBL0001 High Performance Single Board Computer
Technical Product Specification
Hardware Management Overview
WDT #1 can also be configured to take various actions before timing out (for example, SMI, NMI,
nothing) or after timing out (for example, hard reset, power down, or power cycle). In addition, an
event can be logged into the SEL whenever the watchdog timer expires. If WDT #1 expires, the
IPMC is not reset. For more details on the watchdog timer commands and settings, see the IPMI
Specification version 1.5.
On power up, the initial state is that the IPMI WDT #1 is not running. Normally some code (BIOS
or OS level) must send the Reset Watch dog Timer command to start the timer running. The same
code sends a Set W atchdog Timer command first to set up the timer to a known state (see the IPMI
Specification for more details).
When WDT #1 times out, it logs an event into the SEL, provided that the “Don’t Log” flag is false
(see the IPMI 1.5 Specification for details). The SEL event also describes the timeout action taken.
If WDT #1 times out and causes a hard reset, the timer state is equivalent to the power-up state
(that is, not running; either BIOS or the OS must configure and start it). If the host processor is
reset (soft or hard) independent of WDT #1, the firmware disables the watchdog timer.
One of the actions BIOS takes very early in its code is to start the WDT #1 to monitor its boot
progress. When it finishes POST, the BIOS turns off WDT #1 during the OS load period.
WDT #1 parameters are altered according to BIOS control parameters, and WDT #1 is not running
when the OS first (re)starts. The BIOS sets WDT #1 to a length of time longer than the expected
POST time; therefore, BIOS does not actively strobe WDT #1. The flag that determines if a WDT
#1 reset must be hard or soft remains over any type of reset, since it is held in the microcontroller.
3.13.2WDT #2
WDT #2 (implemented in a PLD) must be strobed by the IPMC firmware. If WDT #2 expires, it
isolates the SBC from the backplane IPMB buses and resets the IPMC. There is no method for the
processor to be explicitly notified that the IPMC is reset. Once the IPMC has reset, the main
processors can resume communication with the IPMC. The watchdog timer is set to trigger after 96
seconds, and the IPMC strobes it once a second.
WDT #2 is always running; that is, the counter is always counting. However, a PLD component
controls the IPMC reset and IPMB isolation associated with WDT #2 expiration, ignoring any
WDT event until the IPMC strobes/enables the LTC4300 IPMB in terfaces.
3.13.3WDT #3
WDT #3 is contained within the ICH3 device. This watchdog timer monitors the processor’s first
attempt to fetch an instruction after a power up or hard reset. If the processor has not fetched its
first instruction within the timeout period, the ICH3 resets the processors. Since the processor has
not begun any execution, the ICH3 uses a hard reset.
Intel NetStructure
Technical Product Specification
®
MPCBL0001 High Performance Single Board Computer 63
Hardware Management Overview
3.14LED Status
3.14.1Health LED
The MPCBL0001 SBC supports one bicolor health LED to indicate the SBC’s health status, i.e.,
whether a fault or error condition has been detected on the SBC. This LED is mounted on the front
faceplate and driven by the onboard IPMC. The health LED will only be driven to an error
condition (red) if there is a critical or non-recoverable (major or critical in AdvancedTCA
parlance) condition active on the SBC. Alarms could include exceeding sensor thresholds for
temperature and on-board logic voltages. The health LED remains red until the sensors return to a
normal operating value. Hard-drive failures, boot failures, etc. are not considered critical/major
IPMI states, so the IPMC does not explicitly set the health LED in these cases.
Note: The LED's error state color defaults to red, but the color can be overridden using PICMG 3.0-
Table 26. Health LED
defined commands.
LED Status (right)Meaning
Solid GreenHealthy
Solid Amber/RedFault or error condition
The default color and override capabilities of the LED follow the LED management requirements
defined in Section 3.2.5 of the PICMG 3.0 Specification.
3.14.2OOS (Out Of Service) LED
The MPCBL0001 SBC supports one bicolor “OOS” LED, mounted on the front faceplate. The
LED can be driven to display a red or amber color. When this LED is lit, it indicates that the board
is not in service. Its back-end (payload) power could be OFF or ON. Often the OOS state is entered
when a critical fault occurs on the board. In this state, the back-end (payload) power is turned OFF.
A board could be in this state when its back-end power is OFF but healthy, or when a board is fully
powered but not yet deployed, or during the reset process.
Note: Do not extract a board unless the Hot Swap LED is lit.
Table 27. OOS LED (DS9)
LED Status (left)Meaning
OffIn service
Solid Amber/RedFault or error condition
The default color and override capabilities of the LED follow the LED management requirements
defined in Section 3.2.5 of the PICMG 3.0 Specification.
3.14.3Hot-Swap LED
See Section 3.9, “Hot-Swap Process” on page 51.
64Intel NetStructure
®
MPCBL0001 High Performance Single Board Computer
Technical Product Specification
3.14.4IDE Drive Activity LED
Table 28. IDE Drive Activity LED
LED StatusMeaning
OffNormal/No disk access
Green (Blinking)Disk access (read/write activity)
3.14.5User Programmable LEDs
The MPCBL0001 SBC provides two bicolor LEDs for user-programmable functions. The LEDs
can be driven to display a red, green or amber color. When these LEDs are lit, they indicate a status
of a user-defined function.
Table 29. User Programmable LEDs
LED Status (left)LED Status (right)Meaning
OffOffNo Status
RedRed/GreenActive Status of user defined function
Hardware Management Overview
The user-programmable LEDs are connected to the GPIO pins on the ICH3 device as follows:
Table 30. GPIO Pin Connections
LEDPin
User_Prog_LED1_Red#GPIO21
User_Prog_LED1_GRN#GPIO20
User_Prog_LED2_Red# GPIO28
User_Prog_LED2_GRN#GPIO23
By programming the ICH3 GPIO registers as outputs, then selecting the appropriate state (low for
illumination, high for off), the user enables the LEDs as required. Refer to the ICH3 datasheet in
appendix B for specific GPIO 20, 21, 23, 28 register information.
Intel NetStructure
Technical Product Specification
®
MPCBL0001 High Performance Single Board Computer 65
Hardware Management Overview
3.14.6Network Link/Speed LEDs
The front panel of the SBC provides two LEDs for each Ethernet connection indicating the speed
and link activity for that network connection:
Table 31. Network Link LEDs
For Channel A : L2 / For Channel B : L6
Link LED StatusMeaning
OffNo link
Solid GreenLink established
Blinking GreenLink with activity
NOTE: Refer to Figure 14 and Figure 15 for LED (L2 and L6) placement on the
Table 32. Network Speed LEDs
Front Panel.
For Ethernet controller Channel A : L3 & L4
Speed LED Status
L3L4
Solid YellowOff1 Gbps connection
OffSolid Green100 Mbps connection
OffOff10 Mbps connection
For Ethernet controller Channel B : L7 & L8
Speed LED Status
L7L8
Solid YellowOff1 Gbps connection
OffSolid Green100 Mbps connection
OffOff10 Mbps connection
NOTE: Refer to Figure 14 and Figure 15 for LED (L3, L4, L7 and L8) placement
on the Front Panel.
Meaning
Meaning
3.14.7Ethernet Controller Port State LEDs
The front panel of the SBC provides a bicolor LED for each Ethernet channel that can light to
indicate the Ethernet port state. These LEDs can display a red, green or amber color. The function
of the port state LEDs is user definable. The Ethernet Controller SDP[6:7] GPIO bits for each
channel are the outputs that control the LEDs. SDP[6] is connected to the Green LED, and SPD[7]
is connected to the Red LED.
Refer to the documentation for the Intel
on how to drive these LED signals. Note that existing network drivers may drive these GPIO pins.
66Intel NetStructure
®
82546 Dual Gigabit Ethernet Controller for information
®
MPCBL0001 High Performance Single Board Computer
Technical Product Specification
Table 33. Ethernet Controller Port State LED
LED Status (L1 and L5)Meaning
OffNo Status
Red/Green/AmberActive status of user-defined function
NOTE: Refer to Figure 14 and Figure 15 for LED (L1 and L5) placement on the
Front Panel.
3.14.8Fibre Channel Port State LEDs
The MPCBL0001 SBC supports two Fibre Channel port state LEDs mounted on the front
faceplate. The LEDs are green and yellow. When this LED is lit, it indicates the port state of each
Fibre Channel port. LED states are shown in the table as follows:
Table 34. Fibre Channel Port State LED (DS2, DS3)
Hardware Management Overview
Yellow LED Status (Fibre
Channel 1, left)
ONONPower On
FlashingOFFLoss-of-Sync
ONOFFSignal Acquired
OFFONOn-Line
FLASHFLASHF/W Error
Green LED Status (Fibre
Channel 2, right)
3.15FRU Payload Control
The MPCBL0001 implements the “FRU Control” command as specified in the PICMG 3.0
Specification. Through this command, the payload can be reset, rebooted, or have its diagnostics
initiated.
The FRU payload can be controlled by a command line via the Intel NetStructure
Chassis Management Module (CMM). The following CMM commands are supported by the
MPCBL0001.
Note: The user may issue an RMCP command to control the FRU payload as well. Refer to Table 98 on
page 165 for the associated IPMI command information.
Intel NetStructure
Technical Product Specification
®
MPCBL0001 High Performance Single Board Computer 67
Hardware Management Overview
3.15.1Cold Reset
When this command is initiated, th e board will perform a hard reset as described in Section 3.12.2,
“Hard Reset Request” on page 57.
3.15.2Warm Reset
When this command is initiated, the board will perform a soft reset as described in Section 3.12.3,
“Soft Reset Request” on page 57.
3.15.3Graceful Reboot
This specific payload control command is implemented using system interface messaging
capability and the SMS_ATN bit of the KCS status registers.
The Receive Message Queue is used to hold message data for system software until the system
software can collect it, while the SMS_ATN bit is used to indicate that the IPMC requires attention
from the system software.
The flow diagram below will assist the user who will be developing their system software to
interact with this command.
Figure 11. Flow Diagram for Graceful Reboot Command
OS Agent
IPMCCMM
3
Asserts SMS_ATN signal
Get Message
1. MM sends a frucontrol=2 command to IPMC, initiating a graceful reboot.
2. When the IPMC receives frucontrol=2, it formats a message into the send message queue and
sets the SMS attention flag (SMS_ATN) on the KCS status register.
3. OS Agent polls for SMS_ATN using Get MessageFlags command.
4. OS Agent sends a Get Message command to the IPMC to retrieve the message from the
receive message queue. The Get Message command returns the following data:
4
IPMB Interface KCS Interface
Cmmset –l bladex –d frucontrol –v 2
1
68Intel NetStructure
®
MPCBL0001 High Performance Single Board Computer
Technical Product Specification
Table 36. Returned Values from the Get Message Command
9Data checksum5F2’s complement of the sum of the previous 4 bytes (chk2)
3.15.4Diagnostic Interrupt
The following command provides the capability for an end user to issue a non-maskable interrupt
(NMI) to the payload.
Hardware Management Overview
When issued, the NMI signal to the processor will be asserted. To fully utilize the support of this
command, the user needs to have an NMI handler installed.
1. CMM sends a frucontrol=3 command to IPMC initiating a diagnostic interrupt.
2. When the IPMC receives frucontrol=3, it asserts the NMI signal to the CPU via the GPIO pins
connected to the H_NMI pin.
IPMB Interface GPIO
1
Intel NetStructure
®
MPCBL0001 High Performance Single Board Computer 69
Technical Product Specification
Connectors
Connectors4
Connectors along the rear edge of AdvancedTCA server blades are divided into three distinct
zones, as described in Section 2.3 of the PICMG 3.0 Specification.
• Zone 1 for system management and power distribution
• Zone 2 for data fabric
• Zone 3 for the rear transition module.
As shown in Figure 13, the MPCBL0001 includes several connectors to interface with applicationspecific devices. Some of the connectors are available at the front panel.Each connector is
described briefly in Table 38 on page 73. A detailed description and pinout for each connector is
found in the following sections
MPCBL0001 High Performance Single Board Computer 73
DescriptionDetails
Two 10/100/1000 Ethernet ports
Two 2 Gbit Fibre Channel ports
DescriptionDetails
Connectors
4.1Backplane Connectors
4.1.1Power Distribution Connector (Zone 1)
Zone 1 consists of P10, a 34-pin Positronic header connector that provides the following signals:
• Two -48 VDC power feeds (four signals each; eight signals total)
• Two IPMB ports (two signals each, four signals total)
• Geographic address (eight signals)
• 5.55 Amperes are allocated to MPCBL0001 on the -48 VDC redundant power feeds. This is
equivalent to 200 Watts at the minimum input voltage (-36 VDC). The Zone 1 connector and
pin out is compatible with the backplane for Intel NetStructureTechnical Product Specification.
Note: The analog test and ring voltage pins defined on P10 are left unconnected on MPCBL0001.
The connector used is Positronic part number VPB30W8M6200A1. Figure 16, “Power
Distribution Connector (Zone 1) P10” on page 74 shows the mechanical drawing of the connector.
The pin assignments are given in Figure 39, “Power Distribution Connector (Zone 1) P10 Pin
Assignments” on page 74.
®
MPCHC0001 14U Shelf
Figure 16. Power Distribution Connector (Zone 1) P10
46.10
[1.815] MAX
Table 39. Power Distribution Connector (Zone 1) P10 Pin Assignments
Table 39. Power Distribution Connector (Zone 1) P10 Pin Assignments
8GA3Geographic Addr Bit 325EMI_GNDEMI Chassis Ground
9GA4Geographic Addr Bit 426LOGIC_GNDGnd Ref for Card Logic
10GA5Geographic Addr Bit 527ENABLE_BEnb DC-DC conv, B Feed
11GA6Geographic Addr Bit 628VRTN_A-48 V Return, Feed A
12GA7/PGeo Adr Bit 7 (Odd Parity)29VRTN_B-48 V Return, Feed B
13IPMB_CLK_AIPMB Bus A Clock30- 48 V_EARLY_A-48 V In, Feed A Precharge
14IPMB_DAT_AIPMB Bus A Data31-48 V_EARLY_B-48 V In, Feed B Precharge
15IPMB_CLK_BIPMB Bus B Clock32ENABLE_AEnb DC-DC conv, A Feed
16IPMB_DAT_BIPMB Bus B Data33-48V_A-48 V Input, Feed A
17UnusedNo Connect34-48V_B-48 V Input, Feed B
4.1.2Data Transport Connector (Zon e 2)
Zone 2 consists of one 120-pin HM-Zd connector, labeled P23, with 40 dif feren tial pairs. Th is data
transport connector provides the following signals:
• Two 10/100/1000Base-T/TX Ethernet base fabric channels (four differential signal pairs each,
16 signals total).
• T wo 2 Gbit Fibre Channel ports on the extended fabric (two differential signal pairs each, eight
signals total).
Connectors
The connector used is AMP/Tyco part number 1469001-1, Intel part number A66621-001.
Figure 17, “Data Transport Connector (Zone 2) J23” on page 75 shows a face view of the
connector.
Figure 17. Data Transport Connector (Zone 2) J23
HGFGDGBG
HGFEDCBA
1
2
3
4
5
6
7
8
9
10
The following naming convention describes the signals on this connector. Signal direction is
defined from the perspective of MPCBL0001.
B0899-01
Intel NetStructure
®
MPCBL0001 High Performance Single Board Computer 75
Technical Product Specification
Connectors
P[C]dxp where:
P = Prefix (B=Base Interface [Gigabit Ethernet], F= Fabric Interface [Fibre Channel])
C = Channel (1-2)
d = direction (Tx = Transmit, Rx = Receive)
x = port number (0-1)
Note: A port is two differenti al pairs, one Tx and one Rx
p = polarity (+, -)
The BG, DG, FG and HG (G for Ground) columns contain the ground shields for the four columns
of differential pairs. They have been omitted from the pin out tables below for simplification. All
pins in the BG, DG, FG and HG columns are connected to Logic Ground. The used base fabric
(Gigabit Ethernet) channels are shown in light gray while the used extended fabric (Fibre Channel)
ports appear in dark gray.
Table 40. Data Transport Connector (Zone 2) P23 Pin Assignments
The MPCBL0001 SBC implements the K1 and K2 alignment blocks at the top of Zone 2 and Zone
3, as required in section 2.4.4 of the PICMG 3.0 Specification. These are identified on the
silkscreen as GP1 and GP2. GP1 provides the PICMG 3.0-mandated keying value of 11, and is
either a T yco* 1469373 or a Tyco 1469268 component (or equivalent). GP2 has a solid face and is
used to ensure that RTMs with protruding connectors are not plugged into the MPCBL0001 SBC
or vice versa; the component used for this is either a Tyco 1469374 or a Tyco 1469275-2 (or
equivalent).
76Intel NetStructure
®
MPCBL0001 High Performance Single Board Computer
Technical Product Specification
4.2Front Panel Connectors
4.2.1USB Connector (J12)
MOLEX part Number: 67329-0020
The MPCBL0001 SBC has one vertical USB connector that supports USB 1.1. USB connector JX
is available at the front panel, as shown in Figure 13, “MPCBL0001 SBC Connector Locations” on
page 70. The figure shows its position on the board. See Table 41, “USB Connector (J12) Pin
Assignments” on page 77 for pinout information.
Table 41. USB Connector (J12) Pin Assignments
USB CONNECTORUSB Connector (J12)
Pin #Signal Name
1+5V
2-DATA
3+DATA
4GND
Connectors
4.2.2Serial Port Connector (J17)
A single serial port interface is provided on the front edge of the card using an RJ-45 style shielded
connector. See Figure 13, “MPCBL0001 SBC Connector Locations” on page 70 for its position on
the board. The default connector is an 8-pin RJ-45.
MOLEX Part Number43249-8919
Intel NetStructure
Technical Product Specification
®
MPCBL0001 High Performance Single Board Computer 77
Connectors
Figure 18. Serial Port Connector (J17)
Optional Top Ground Tabs
Shielded Modular Jack Assembly
molex
Optional Side
Ground tabs
(2 places)
.724 REF.
18.39t
.512
13.00
Table 42. Serial Port Connector (J17) Pin Assignments
Connector Pin
Number
1RTS
2DTR
3TXD
4GND
5GND
6RXD
7DSR
8CTS
Serial Port Signal
.120
3.05
REF.
(outside)
REF.
.120
3.05
REF.
.128
3.25
.427
10.85
.829
21.05
.120
3.05
B0902-01
78Intel NetStructure
®
MPCBL0001 High Performance Single Board Computer
Technical Product Specification
Figure 19. DB9 to RJ-45 Pin Translation
Connectors
Intel NetStructure
Technical Product Specification
®
MPCBL0001 High Performance Single Board Computer 79
Connectors
4.2.3Fibre Channel Small Form-Factor Pluggable (SFP)
Receptacle (J34 and J35)
AMP part number: 1367073-1
The MPCBL0001 SBC has two SFP receptacles that support either the copper or fiber module
interface. Fibre Channel connector J34 and J35 are available at the front panel. See Figure 13,
“MPCBL0001 SBC Connector Locations” on page 70 for its position on the board. See Table 44,
“Fibre Channel SFP Pin Assignments” on page 81 for pinout informat ion.
Refer to the Intel NetStructure® MPCBL0001 Compatibility Report for a list of SFP optical
transceivers that have been validated. The report can be downloaded from
There are three 64-pin connectors that make up the PMC card connection:
MOLEX Part Number: 71439-0864
These connectors and pinouts are defined by the following industry standard specifications:
• Draft Standard Physical and Environmental Layers for PCI Mezzanine Cards: PMC IEEE
(MMSC) P1386.1/Draft 2.3, October 9, 2000
• Draft Standard for a Common Mezzanine Card Family: CMC IEEE (MMSC) P1386/Draft 2.3,
October 9, 2000
The PMC slot is available at the front panel. See Figure 13, “MPCBL0001 SBC Connector
Locations” on page 70 for their positions on the board. Pin assignments are listed in Table 45,
“PMC Connector Pin Assignments - 32 Bit” on page 82 and Table 46, “PMC Connector Pin
Assignments - 64 Bit” on page 83.
Intel NetStructure
Technical Product Specification
®
MPCBL0001 High Performance Single Board Computer 81
Connectors
Table 45. PMC Connector Pin Assignments - 32 Bit
J2532 Bit PCIJ2632 Bit PCI
Pin SignalSignalPinPin SignalSignalPin
1TCK-12V21+12VTRST#2
3GroundINTA#43TMSTDO4
5INTB#INTC#65TDIGround6
7BUSMODE1#+5 V87Ground
9INTD#PCI-RSVD109PCI-RSVDPCI-RSVD10
11Ground(n/c) 3.3 Vaux 1211BUSMODE2#+3.3 V12
13CLKGround1413RST#BUSMODE3#14
15GroundGNT[0]#1615+3.3 VBUSMODE4#16
17REQ[0]#+5 V1817PME#Ground18
19+3.3V (V I/O)AD[31]2019AD[30]AD[29]20
21AD[28]AD[27]2221GroundAD[26]22
23AD[25]Ground2423AD[24]+3.3 V24
25GroundC/BE[3]#2625IDSEL (AD17)AD[23]26
27AD[22]AD[21]2827+3.3 VAD[20]28
29AD[19]+5 V3029AD[18]Ground30
31+3.3V (V I/O)AD[17]3231AD[16]C/BE[2]#32
33FRAME#Ground3433Ground
35GroundIRDY#3635TRDY#+3.3 V36
37DEVSEL#+5 V3837GroundSTOP#38
39GroundLOCK#4039PERR#Ground40
PCI-RSVDPCI-RSVD4241+3.3VSERR#42
41
43PARGround4443C/BE[1]#Ground44
45+3.3 V (V I/O)AD[15]4645AD[14]AD[13]46
47AD[12]AD[11]4847M66ENAD[10]48
49AD[09]+5 V5049AD[08]+3.3 V50
51GroundC/BE[0]#5251AD[07]
53AD[06]AD[05]5453+3.3VPMC-RSVD54
55AD[04]Ground5655PMC-RSVDGround56
57+3.3 V (V I/O)AD[03]5857PMC-RSVDPMC-RSVD58
59AD[02]AD[01]6059GroundPMC-RSVD60
61AD[00]+5 V6261ACK64#+3.3 V62
63GroundREQ64#6463Ground
PCI-RSVD8
PMC-RSVD34
PMC-RSVD52
PMC-RSVD64
82Intel NetStructure
®
MPCBL0001 High Performance Single Board Computer
Technical Product Specification
Table 46. PMC Connector Pin Assignments - 64 Bit
J2764 Bit PCI
Pin SignalSignalPin
PCI-RSVDGround2
1
3GroundC/BE[7]#4
5C/BE[6]#C/BE[5]#6
7C/BE[4]#Ground8
9+3.3 V (V I/O) PAR6410
11AD[63]AD[62]12
13AD[61]Ground14
15GroundAD[60]16
17AD[59]AD[58]18
19AD[57]Ground20
21+3.3 V (V I/O) AD[56]22
23AD[55]AD[54]24
25AD[53]Ground26
27GroundAD[52]28
29AD[51]AD[50]30
31AD[49]Ground32
33GroundAD[48]34
35AD[47]AD[46]36
37AD[45]Ground38
39+3.3 V (V I/O) AD[44]40
41AD[43]AD[42]42
43AD[41]Ground44
45GroundAD[40]46
47AD[39]AD[38]48
49AD[37]Ground50
51GroundAD[36]52
53AD[35]AD[34]54
55AD[33]Ground56
57+3.3 V (V I/O) AD[32]58
PCI-RSVDPCI-RSVD60
59
61PCI-RSVDGround62
63GroundPCI-RSVD64
Connectors
Intel NetStructure
®
MPCBL0001 High Performance Single Board Computer 83
I/O Address:0x0CF8 Accessed as a Dword
Default Value:0x00000000
Access:Read/Write
Size:32 bits
CONFIG_ADDRESS is a 32-bit I/O register that can be accessed only as a Dword. A byte or word
reference passes through the Configuration Address Register and hub link interface HI_A onto the
PCI_A bus as an I/O cycle. The CONFIG_ADDRESS register contains the Bus Number, Device
Number, Function Number , and Register Number for which a subsequent PCI configuration access
is intended. This register is defined by the PCI Bus Specification.
Table 48. Configuration Address Register Bit Assignments
Bit3130 2423 1615 1110 87 21 0
0R0000R Default
BitDescription
31
30:24Reserved (These bits are read only and have a value of 0).
23:16Bus Number: Contains the bus number being targeted by the configuration cycle.
15:11Device Number: Selects one of the 32 possible devices per bus.
10:8Function Number: Selects one of eight possible functions within a device.
7:2
1:0Reserved.
Configuration Enable (CFGE): When this bit is set to 1, accesses to the PCI configuration space
are enabled. When this bit is reset to 0, accesses to the PCI configuration space are disabled.
Register Number: This field selects one register within a particular bus, device, and function as
specified by the other fields in the Configuration Address Register. This field is mapped to A[7:2]
during HI_A-D configuration cycles.
5.1.2Configuration Data Register MCH CONFIG_ADDRESS
MPCBL0001 High Performance Single Board Computer 85
Addressing
CONFIG_DATA is a 32-bit read/write window into the PCI configuration space. The portion of
configuration space that is referenced by CONFIG_DATA is determined by the contents of
CONFIG_ADDRESS.
Table 49. Configuration Data Register Bit Assignments
BitDescription
Configuration Data Window (CDW): If bit 31 of CONFIG_ADDRESS is set to 1, any I/O access to
31:0
the CONFIG_DATA register is mapped to configuration space pointed to by the contents of
CONFIG_ADDRESS.
5.2I/O Address Assignments
I/O port addresses are divided among the on-board devices.These devices include:
• ICH3
• ISP2312 Fibre Channel controller
• 82546 Ethernet controller
• SMSC LPC47B272 SIO
• MCH
• IPMC
Please refer to the respective device specifications for specific I/O address usage.
The MCH uses only I/O ports 0xCF8 and 0xCFC for PCI configuration cycle generation. These
registers were shown in Section 5.1.1 and Section 5.1.2. The P64H2 forwards appli cable I/O
transactions to its attached PCI buses. The ISP2312 may be programmed to map its 256-byte bank
of registers to memory and/or I/O space.
Table 50 lists document references to I/O descriptions. Pl ease refer to Appendix A, “Reference
Documents” for a list of the referenced documents and their complete titles, revisions, and
NOTE: The OS may need to be recompiled to support memory above 4 Gbytes.
The Firmware Hub(s) also appears at the aliased address of (4 Gbyte – 4 Mbyte).
The MCH provides the capability to reclaim the physical memory overlapped by memory-mapped
I/O devices, BIOS, and I/O APICs that reside ju st below 4 Gbytes. This memory may be remapped
to physical memory at the address defined by the TOLM register.
Intel NetStructure
Technical Product Specification
®
MPCBL0001 High Performance Single Board Computer 87
Addressing
5.4IPMC Addresses
The IPMC supports 6 I2C/SMB buses. IPMC buses 0 and 1 provide redundant IPMB connections.
The ADM1026 device is connected to SMBus 3 and provides voltage measurement capability and
additional board configuration status.
Table 52. SMBus Addresses
8-bit Address (W/R)SMBusDescription
5C/5D3ADM1026
A8/A93SEL EEPROM
88Intel NetStructure
®
MPCBL0001 High Performance Single Board Computer
Technical Product Specification
Specifications
Specifications6
This chapter defines the MPCBL0001 operating and nonoperating environments. It also documents
the procedures followed to determine the reliability of MPCBL0001.
6.1Mechanical Specifications
6.1.1Board Outline
Figure 20 and Figure 21 are annotated illustrations of the MPCBL0001 SBC showing the locations
of major components. The board dimensions are 280 mm x 322.25 mm. The board pitch is 1.2”
(30.48 mm).
Figure 20 is applicable to the following SKU and TA numbers:
• MPCBL0001F04 with TA number C55360-011 or below.
• MPCBL0001N04 with TA number C13354-010 or below.
Figure 21 is applicable to the following SKU and TA numbers:
• MPCBL0001F04 with TA number C55360-014 or below.
• MPCBL0001N04 with TA number C13354-013 or below.
Intel NetStructure
Technical Product Specification
®
MPCBL0001 High Performance Single Board Computer 89
A - IDE connector (J24)
B - Power mezzanine card + cover
C - PMC connectors
D - Barcode: serial number + PBA version number
E - DIMM banks
F - EMI filter mezzanine
G - Barcode: MAC Address 1 (Note: MAC Address 2 is an incremental value of MAC Address 1. This MAC
Address labeling applies to boards with TA# C13354-015 and C55360-016 (or below).)
Intel NetStructure
®
MPCBL0001 High Performance Single Board Computer 91
Technical Product Specification
Specifications
6.1.2Backing Plate
The MPCBL0001 SBC has a rugged metal backing plate that forms a single-piece face plate. This
backing plate is made of 1.2 mm (0.048") steel which has been zinc post-plated to resist corrosion
and rust. The solid backing plate provides PCB stiffening, enhanced EMI protection from adjacent
boards, and protection during flame tests. The backing plate improves serviceability by making the
SBC more durable.
Four holes are provided in the bottom of the backing plate for mounting an optional hard drive in
the provided hard drive carrier (with the included M3 screws). Four additional holes are provided
for securing an optional PMC through the front or rear standard mounting positions.
Caution: Removing the backing plate can damage the components on the board and may void the warranty.
No user-serviceable parts are available under the PCB. Do not remove the face plate/backing plate.
6.1.3Component Height
Figure 22 on page 93 and Figure 23 on page 94 detail maximum component heights on both the
primary and secondary sides of the MPCBL0001 SBC.
92Intel NetStructure
®
MPCBL0001 High Performance Single Board Computer
Technical Product Specification
Specifications
Figure 22. MPCBL0001 SBC Front Panel Dimensions – FC SKU (PMC and Connectors)
Intel NetStructure
Technical Product Specification
®
MPCBL0001 High Performance Single Board Computer 93
Specifications
Figure 23. MPCBL0001 SBC Front Panel Dimensions – FC SKU (Screws and LEDs)
94Intel NetStructure
®
MPCBL0001 High Performance Single Board Computer
Technical Product Specification
Specifications
Figure 24. MPCBL0001 SBC Front Panel Dimensions – Non FC SKU (PMC and Connectors)
Intel NetStructure
Technical Product Specification
®
MPCBL0001 High Performance Single Board Computer 95
Specifications
Figure 25. MPCBL0001 SBC Front Panel Dimensions – Non-FC SKU (Screws and LED)
96Intel NetStructure
®
MPCBL0001 High Performance Single Board Computer
Technical Product Specification
6.2Environmental Specifications
The MPCBL0001 SBC meets the board-level specifications as specified in the Intel Environmental
Standards Handbook – Telco Specification Document No. A78805-01. The test methodology is a
combination of Intel and NEBs test requirements with the intent that the product will pass pure
system-level NEBs testing. Intel will not be completing NEBs testing on the SBC. The following
table summarizes environmental limits, both operating and nonoperating.
Table 53. Environmental Specifications
ParameterConditionsDetailed Specification
Temperature
(Ambient)
AirflowOperating300 linear feet per minute (LFM) minimum
6.3.1Mean Time Between Failure (MTBF) Specifications
Calculation Type:MTBF/FIT Rate
Standard:Telcordia Standard SR-332 Issue 1
Methods:Method I, Case I, Quality Level II
The calculation results were generated using the references and assumptions listed. This report and
its associated calculations supersede all other released MTBF and Failure in Time (FIT)
calculations of earlier report dates. The reported failure rates do not represent catastrophic failure.
Catastrophic failure rates will vary based on application environment and features critical to the
intended function.
Note: Incorporating an optional IDE Hard-disk Drive (HDD) will significantly impact the Reliability
Specifications.
Table 54. Reliability Estimate Data
9
Failure Rate (FIT)
MTBF
8,000Failures in 10
125,000Hours
hours
Intel NetStructure
Technical Product Specification
®
MPCBL0001 High Performance Single Board Computer 97
Specifications
6.3.1.1Environmental Assumptions
Failure rates are based on a 40° C ambient temperature.
• Ground, fixed, controlled environment with an environmental adjustment factor equal to 1.0.
6.3.1.2General Assumptions
• Component failure rates are constant.
• Board-to-system interconnects included within estimates.
• Non-electrical components (screws, mechanical latches, labels, covers, etc.) are not included
within estimations.
• Printed circuit board is considered to have a 0 FIT rate.
6.3.1.3General Notes
• Method I, Case I = Based on parts count. Equipment failure is estimated by totaling device
failures rates and quantities used.
• Quality Level II = Devices purchased to specifications, qualified devices, vendor lot-to-lot
controls for AQLs and DPMs.
• Where available, direct component supplier predictions or actual FIT rates have been used.
• The SBC MTBF does not include addition of the 2.5” HDD. The product MTBF could be
significantly impacted by adding a HDD. Please contact the HDD manufacturer for specific
component and relevant operational MTBF information.
6.3.2Power Consumption
The power consumed by the Intel NetStructure® MPCBL0001 High Performance Single Board
Computer SBC is dependent on the type and speed of processors used and the amount of memory
installed. Table 55, “Total Measured Power” on page 98 is based on the use of two Low Volt age
®
Intel
Xeon™ processors. Typical values were obtained by running the Windows* 2000*-based
application “Drive Reaper” against networked shared drives. “Max” values were obtained by
running Intel's DOS* based “Maxpower” utility, version 6.0.
Note: A TriEMS card was installed for all power management tests. The TriEMS dissipates 550 mW
typical.
Note: A TriEMS card was installed for all power management tests. The TriEMS dissipates 550 mW
typical. The power level that was sent to the shelf manager at M3 state is 152W . 10W margin from
the maximum power was an allowance for different model of PMC and DIMMs module.
Table 55. Total Measured Power
MemoryDual 2.0 GHz (400 Mhz FSB)
8GByte (Four 2-GByte DIMMs)Typical power = 127 W
Max power = 142 W
98Intel NetStructure
®
MPCBL0001 High Performance Single Board Computer
Technical Product Specification
6.3.3Cooling Requirements
The Intel NetStructure® MPCBL0001 High Performance Single Board Computer SBC should be
installed vertically in a chassis, with bottom-to-top airflow. Airflow is expected to be evenly
distributed across the bottom edge of the installed MPCBL0001 blade and maintain at least 300
LFM average airflow.
Most components on the MPCBL0001 blade are specified to operate with a localized ambient
temperature up to 70° C and do not require heat sinks. The MPCBL0001 blade uses two custom
heat sinks, one per processor (see Figure 26, “Low Voltage Intel
page 99.) The rate of airflow specified above is critical to ensuring that the blade operates as
designed.
®
Figure 26. Low Voltage Intel
Xeon™ Processor Heatsink
Specifications
®
Xeon™ Processor Heatsink” on
6.4Board Layer Specifications
Material: TG180 FR4
Layers: 14
Copper:
• Outer layers 1 and 14 are 1 oz copper
• Middle planes 7 & 8 are 2 oz copper
• All others are 1 oz copper.
6.5Weight
The weight of the baseboard (N04 and F04) is 3.0645 kg (6.75 lbs.) without any packaging
materials.
B0903-01
Intel NetStructure
Technical Product Specification
®
MPCBL0001 High Performance Single Board Computer 99
BIOS Features
BIOS Features7
7.1Introduction
The Intel NetStructure® MPCBL0001 High Performance Single Board Computer SBC uses an
Intel/AMI BIOS, which is stored in flash memory and updated using a disk-based program. In
addition to the BIOS and BIOS setup program, the flash memory contains POST and Plug and Play
support.
The BIOS displays a message during POST identifying the type of BIOS and a revision code. Refer
to the specification update for the latest default settings.
7.2BIOS Flash Memory Organization
MPCBL0001 contains two Firm Ware Hub (FWH) devices (see Figure 1, “Intel NetStructure®
MPCBL0001 SBC Block Diagram” on page 15). The first one is the Primary FWH, which holds
the BIOS code that executes during POST. The second is the Backup FWH, which recovers the
system when the Primary FWH is corrupted. The N82802AC FWH includes an 8 Mbit (1024
KByte) symmetrical flash memory device. Internally, the device is grou ped into sixteen 64-KByte
blocks that are individually erasable, lockable, and unlockable.
7.3Complementary Metal-Oxide Semiconductor
(CMOS)
CMOS RAM is a nonvolatile storage that stores data needed by the BIOS. The data consists of
certain onboard configurable settings, including time and date. CMOS resides in the ICH3 and is
powered by the Supercap when the blade is power off. The settings in the BIOS setup menu are
stored in the CMOS RAM and are often called CMOS settings.
7.3.1Copying and Saving CMOS Settings
The BIOS/CMOS flash update utility (flashlnx or flashdos) loads a fresh copy of the BIOS into
flash ROM. It has the capability to save the CMOS settings from the MPCBL0001 SBC. The
CMOS settings file can be copied to a file. This file can be saved in a directory specified by the
user. The filename also can be specified by user, such as CMOS.BIN.
With the BIOS/CMOS flash utility and CMOS.bin file, user is able to copy CMOS settings to
another MPCBL0001 SBC, thus minimizing the effort to reconfigure the preferred CMOS settings
across all boards.
This BIOS/CMOS flash utility that is designed to run under MontaVista* Carrier Grade Linux* 3.0
should be on the local hard disk of the MPCBL0001. Any user who is able to communicate with
the MPCBL0001 via Telnet would be able to execute to copy and save the CMOS remotely.
100Intel NetStructure
®
MPCBL0001 High Performance Single Board Computer
Technical Product Specification
Loading...
+ hidden pages
You need points to download manuals.
1 point = 1 manual.
You can buy points or you can get point for every manual you upload.