Table 2. 6-Pin PCI Express Power Connector Pinout ...................................... 10
Table 3. 8-Pin PCI Express Power Connector Pinout ...................................... 10
Table 4. Configuration with External PCI Express Connectors ........................... 11
Table 5. Languages Supported ............................................................... 13
Tesla K10 GPU Accelerator BD-06280-001_v05 | iv
OVERVIEW
The NVIDIA® Tesla® K10 graphics processing unit (GPU) accelerator is a PCI Express,
double-wide, full height (4.376 inches by 10.5 inches by 1.52 inches) form factor
computing module comprising two NVIDIA GK104 GPUs. The Tesla K10 offers a total
of 8 GB of GDDR5 on-board memory (4 GB per GPU) and supports PCI Express Gen3.
The Tesla K10 can be configured by the OEM or by the end user to enable or disable
ECC or error correcting codes that can fix single-bit errors and detect double-bit errors.
Enabling ECC will cause some of the memory to be used for the ECC bits, so the user
available memory will decrease by 10%. On the Tesla K10 the ECC protection is for
DRAM only.
KEY FEATURES
GPU
The Tesla K10 GPU accelerator has two GK104 GPUs. Characteristics for both GPUs are
as follows:
Number of processor cores: 1536 per GPU
Processor core clock: 745 MHz
Package size: 40 mm × 40 mm 1745-pin ball grid array (BGA)
One 6-pin PCI Express power connector
One 8-pin PCI Express power connector
Memory
Memory clock: 2.5 GHz
Interface: 256-bit
● Total board memory: 8 GB (4 GB per GPU)
● 32 pieces of 128M × 16 GDDR5, SDRAM (per GPU)
BIOS
2Mbit Serial ROM
Overview
TESLA K10 BLOCK DIAGRAM
Figure 1 is the block diagram for Tesla K10 GPU accelerator. It comprises two identical
GK104 GPUs, each with 4 GB of GDDR5 memory. The GPUs are connected via a PCI
Express switch. The board supports PCI Express Gen3.
Figure 1. Tesla K10 Block Diagram
Tesla K10 GPU Accelerator BD-06280-001_v05 | 2
Overview
900-22055-0010-000: Airflow towards the bracket
CONFIGURATION
The Tesla K10 boards are available in the following configurations (Table 1) based on the
orientation of the airflow inside the system.
Table 1. Board Configurations
Specifications Tesla K10
Generic SKU reference
Chip 2× GK104
Package size GPU 42.5 mm × 42.5 mm 1981-pin ball grid array (BGA)
Processor clock 745 MHz
Memory clock 2.5 GHz
Memory size 4 GB per GPU (8 GB per boa rd)
Memory I/O 256-bit GDDR5
Memory configuration 32 pieces of 128M × 16 GDDR5 SDRAM
External connectors None
Internal connectors and headers
Board power 225 W
•
• 900-22055-0020-000: Airflow away from the bracket
• 8-pin PCI Express power connector
• 6-pin PCI Express power connector
Tesla K10 GPU Accelerator BD-06280-001_v05 | 3
MECHANICAL SPECIFICATIONS
PCI EXPRESS SYSTEM
The Tesla K10 boards (Figure 2) conform to the PCI Express full height (4.376 inches by
10.5 inches) form factor. Figure 2 is shown without the bracket.
4.376 inches
Figure 2. Tesla K10 GPU Accelerator
10.5 inches
Note: The final product will ship with an 8 and 6-pin connector and shroud.
Tesla K10 GPU Accelerator BD-06280-001_v05 | 4
Mechanical Specifications
STANDARD I/O CONNECTOR PLACEMENT
As shown in Figure 3, the Tesla K10 includes a vented bracket. If you are an OEM who
qualifies for bracket modifications, you have the option of receiving your modules with
no bracket installed.
Figure 3. Tesla K10 Bracket
Tesla K10 GPU Accelerator BD-06280-001_v05 | 5
Mechanical Specifications
If you need to remove the standard bracket follow these simple steps:
1. Remove the two shoulder screws on the back side of the PCB.
2. Remove the two flat head screws on the bracket exhaust face.
3. Remove the bracket.
4. Slide the washer in between the PCB and the backplate to maintain clearance
between the PCB and the backplate.
5. Attach the shoulder screws.
Figure 4. Tesla K10 with Bracket
Figure 5. Tesla K10 without Bracket
Tesla K10 GPU Accelerator BD-06280-001_v05 | 6
Mechanical Specifications
INTERNAL CONNECTORS AND HEADERS
The Tesla K10 GPU accelerator supports the following internal connectors and headers.
8-pin PCI Express power connector (can be used with a 6-pin power cable)
6-pin PCI Express power connector
External PCI Express Power Connectors
The Tesla K10 GPU accelerator is a performance optimized, high-end product and uses
power from the PCI Express connector as well as external power connectors. The board
can be used in two different ways.
One 8-pin PCI Express power connector or
Two 6-pin PCI Express power connectors
Figure 6 and Figure 7 show the specifications and Table 2 and Table 3 show the pinouts
for the 6-pin and 8-pin PCI Express power connectors.
Tesla K10 GPU Accelerator BD-06280-001_v05 | 7
Mechanical Specifications
Figure 6. 6-Pin PCI Express Power Connector
Tesla K10 GPU Accelerator BD-06280-001_v05 | 8
Mechanical Specifications
Figure 7. 8-Pin PCI Express Power Connector
Tesla K10 GPU Accelerator BD-06280-001_v05 | 9
Table 2. 6-Pin PCI Express Power Connector Pinout
Pin Number Description
1 +12 V
2 +12 V
3 +12 V
4 GND
5 Sense
6 GND
Table 3. 8-Pin PCI Express Power Connector Pinout
Pin Number Description
1 +12 V
2 +12 V
3 +12 V
4 Sense1
5 GND
6 Sense0
7 GND
8 GND
Mechanical Specifications
Tesla K10 GPU Accelerator BD-06280-001_v05 | 10
POWER SPECIFICATIONS
The Tesla K10 GPU accelerator requires power from the PCI Express connector as well
as one or two auxiliary power connectors.
Table 4. Configuration with External PCI Express Connectors
Connector Type 6-Pin Power Connector Supported Notes
8-pin connected 6-pin connected Yes
8-pin connected No cable installed Yes 8-pin cable must supply 175
W
6-pin connected N/A No 6-pin cable in the 8-pin
connector is not supported.
Not installed N/A No 8-pin connector should
always be connected.
Note: Detailed information about power draw by rail is available to authorized
system partners in the Tesla K10 Board System Design Guide (DG-06105-001)
Tesla K10 GPU Accelerator BD-06280-001_v05 | 11
SUPPORT INFORMATION
CERTIFICATES AND AGENCIES
Agencies
Australian Communications Authority and Radio Spectrum Management Group of
New Zealand (C-Tick)
Bureau of Standards, Metrology, and Inspection (BSMI)
Conformité Européenne (CE)
Federal Communications Commission (FCC)
Industry Canada - Interference-Causing Equipment Standard (ICES)
Korean Communications Commission (KCC)
Underwriters Laboratories (cUL)
Voluntary Control Council for Interference (VCCI)
Tesla K10 GPU Accelerator BD-06280-001_v05 | 12
LANGUAGES
Table 5. Languages Supported
Support Information
Windows Server
2008 and Windows
English (US) X X
English (UK) X
Arabic X
Chinese, Simplified X
Chinese, Traditional X
Danish X
Dutch X
Finnish X
French X
French (Canada) X
German X
Italian X
Japanese X
Korean X
Norwegian x
Portuguese (Brazil ) X
Russian X
Spanish X
Spanish (Latin America)
Swedish X
Thai X
Server 2008 R2
Linux
Note: NVIDIA’s CUDA™ sof tware is only supported in E n glish (U.S.)
Tesla K10 GPU Accelerator BD-06280-001_v05 | 13
Notice
ALL NVIDIA DESIGN SPECIFICATIONS, REFERENCE BOARDS, FILES, DRAWINGS, DIAGNOSTICS, LISTS, AND OTHER
DOCUMENTS (TOGETHER AND SEPARATELY, “MATERIALS”) ARE BEING PROVIDED “AS IS.” NVIDIA MAKES NO
WARRANTIES, EXPRESSED, IMPLIED, STATUTORY, OR OTHERWISE WITH RESPECT TO THE MATERIALS, AND
EXPRESSLY DISCLAIMS ALL IMPLIED WARRANTIES OF NONINFRINGEMENT, MERCHANTABILITY, AND FITNESS FOR
A PARTICULAR PURPOSE.
Information furnished is believed to be accurate and reliable. However, NVIDIA Corporation assumes no
responsibility for the consequences of use of such information or for any infringement of patents or other
rights of third parties that may result from its use. No license is granted by implication of otherwise under
any patent rights of NVIDIA Corporation. Specifications mentioned in this publication are subject to change
without notice. This publication supersedes and replaces all other information previously supplied. NVIDIA
Corporation products are not authorized as critical components in life support devices or systems without
express written appro v al of NVIDIA Corporati on.
HDMI
HDMI, the HDMI logo, and High-Definition Multimedia Interface are trademarks or registered trademarks of
HDMI Licensing LLC.
ROVI Compliance Statement
NVIDIA Products that support Rovi Corporation’s Revision 7.1.L1 Anti-Copy Process (ACP) encoding technology
can only be sold or distributed to buyers with a valid and existing authorization from ROVI to purchase and
incorpora te the device into buyer’s products.
This device is protected by U.S. patent numbers 6,516,132; 5,583,936; 6,836,549; 7,050,698; and 7,492,896
and other intellectual property rights. The use of ROVI Corporation's copy protection technology in the
device must be authorized by ROVI Corporation and is intended for home and other limited pay-per-view use s
only, unless otherwise authorized in writing by ROVI Corporation. Reverse engineering or disassembly is
prohibited.
OpenCL
OpenCL is a trademark o f App le I nc. us ed under license to the Khronos Group Inc.
Trademarks
NVIDIA, the NVIDIA logo, CUDA, and Tesla are trademarks and/or registered trademarks of N VIDIA Corporation
in the U.S. and other countries. Other company and product names may be trademarks of the respective
companies with which they are associated.