Product Brief
TESLA P100 PCIe
GPU ACCELERATOR
PB-08248-001_v01 | October 2016
DOCUMENT CHANGE HISTORY
PB-08248-001_v01
Version Date Authors Description of Change
01 October 3, 2016 GG, SM Initial Release
Tesla P100 PCIe
GPU Accelerator PB-08248-001_v01 | ii
TABLE OF CONTENTS
Overview ........................................................................................... 1
Specifications ...................................................................................... 2
Product Specifications ......................................................................... 2
Thermal Specifications ......................................................................... 4
System Airflow Requirements .............................................................. 5
Power and Electrical Specifications .......................................................... 6
Design Discussion ................................................................................. 7
Form Factor ...................................................................................... 7
Power Connector Placement .................................................................. 8
CPU 8-Pin to PCIe 8-Pin Dongle ............................................................ 9
Extenders ........................................................................................ 9
Support Information ............................................................................. 11
Certificates and Agencies ..................................................................... 11
Certifications ................................................................................ 11
Agencies ...................................................................................... 11
Languages ....................................................................................... 12
Tesla P100 PCIe
GPU Accelerator PB-08248-001_v01 | iii
LIST OF FIGURES
Figure 1. Tesla P100 PCIe Airflow Directions with Optional I/O Bracket ................ 5
Figure 2. Tesla P100 PCIe Board Dimensions with Optional I/O Bracket ................. 7
Figure 3. CPU 8-Pin Power Connector with Optional I/O Bracket ........................ 8
Figure 4. CPU 8-Pin to PCIe 8-Pin Dongle..................................................... 9
Figure 5. Long Offset Extender ............................................................... 10
Figure 6. Straight Extender .................................................................... 10
LIST OF TABLES
Table 1. Product Specifications ............................................................... 2
Table 2. Memory Specifications ............................................................... 3
Table 3. Software Specifications.............................................................. 3
Table 4. Board Environment and Reliability Specifications ............................... 4
Table 5. Thermal Specifications .............................................................. 4
Table 6. Minimum CFM Requirements per Inlet Air Temperature ........................ 6
Table 7. Input EDP Continuous Specifications .............................................. 6
Table 9. Supported Auxiliary Power Connections .......................................... 8
Table 10. Languages Supported ............................................................... 12
Tesla P100 PCIe
GPU Accelerator PB-08248-001_v01 | iv
OVERVIEW
The NVIDIA® Tesla® P100 GPU Accelerator for PCIe is a dual-slot 10.5 inch PCI Express
®
Gen3 card with a single NVIDIA
a passive heat sink for cooling, which requires system air flow to properly operate the
card within its thermal limits. The Tesla P100 PCIe supports double precision (FP64),
single precision (FP32) and half precision (FP16) compute tasks, unified virtual memory
and page migration engine.
Pascal™ GP100 graphics processing unit (GPU). It uses
™
For performance optimization, NVIDIA GPU Boost
feature is supported. By adjusting
the GPU clock dynamically, maximum performance is achieved within the power cap
limit.
Tesla P100 PCIe boards are shipped with ECC enabled by default to protect the GPU’s
memory interface and the on-board memories. ECC protects the memory interface by
detecting any single, double, and all odd-bit errors. The GPU will replay any memory
transaction that have an ECC error until the data transfer is error-free. ECC protects the
DRAM content by fixing any single-bit errors and detecting double-bit errors. There is
no replay associated with ECC. The Tesla P100 PCIe with HBM2 memory has native
support for ECC and has no ECC overhead, both in memory capacity and bandwidth.
For more information on compute capabilities, HBM2, unified virtual memory, and page
migration engine visit NVIDIA official website.
The Tesla P100 for PCIe is available in two memory configurations:
Tesla P100 PCIe with 16GB HBM2
Tesla P100 PCIe with 12GB HBM2
As a board design, both the memory configurations are identical and this product brief
covers both configurations.
Tesla P100 PCIe
GPU Accelerator PB-08248-001_v01 | 1
SPECIFICATIONS
PRODUCT SPECIFICATIONS
Table 1 provides the product specifications for the Tesla P100 PCIe GPU Accelerator.
Table 1. Product Specifications
Specification Tesla P100 PCIe 16GB Tesla P100 PCIe 12GB
Product SKUs NVPN: 699-2H400-0201-XXX NVPN: 699-2H400-0202-XXX
Total board power 250 W
GPU SKUs GP100-892-A1 GP100-893-A1
PCI Device IDs Device ID: 0x15F8
Vendor ID: 0x10DE
Sub-Vendor ID: 0x10DE
Sub-System ID: 0x118F
NVIDIA® CUDA® cores 3584
GPU
clocks
VBIOS
PCI Express interfaces PCI Express 3.0 ×16
Power connectors and headers One CPU 8-pin auxiliary power connector
Weight
Base 1189 MHz
Maximum boost 1328 MHz
EEPROM size 4 Mbit
UEFI Supported
Lane and polarity reversal supported
Board 1177 Grams
Bracket with screws 21 Grams
Long offset extender 52 Grams
Straight extender 42 Grams
Device ID: 0x15F7
Vendor ID: 0x10DE
Sub-Vendor ID: 0x10DE
Sub-System ID: 0x11DA
Tesla P100 PCIe
GPU Accelerator PB-08248-001_v01 | 2
Specifications
Table 2 provides the memory specifications for the Tesla P100 PCIe board.
Table 2. Memory Specifications
Specification Tesla PCIe 16GB Tesla PCIe 12GB
Maximum memory clock 715 MHz 715 MHz
Memory size 16GB HBM2 12GB HBM2
Memory bus width 4096-bit 3072-bit
Peak memory bandwidth Up to 732 GB/s Up to 549 GB/s
Table 3 provides the software specifications that apply to both the Tesla P100 PCIe 16GB
and the Tesla P100 PCIe 12GB.
Table 3. Software Specifications
Specification Description
Compatibility mode supported Compute (default)
Graphics
Base address BAR0: 16 MB
BAR1: 16 GB (Compute) or 256 MB (Graphics)
BAR3: 32 MB
I/O BAR: 4 KB (Graphics only)
PCI class code 0x03 - Display Controller
PCI sub-class code 0x00 - VGA-Compatible Controller (Graphics)
0x02 - 3D Controller (Compute)
ECC support Supported (Enabled by default)
SMBus (8-bit address) 0x9E (write), 0x9F (read)
SMBus direct access Supported
SMBPBI (SMBus Post Box Interface) Supported
Tesla P100 PCIe
GPU Accelerator PB-08248-001_v01 | 3
Specifications
Table 4 provides the environment conditions specifications to both the Tesla P100 PCIe
16GB and the Tesla P100 PCIe 12GB.
Table 4. Board Environment and Reliability Specifications
Specification Condition
Operating temperature 0 °C to 45 °C
Storage temperature -40 °C to 75 °C
Operating humidity 5% to 90% relative humidity
Storage humidity 5% to 95% relative humidity
Mean time between failures (MTBF)
Uncontrolled environment: 769566 hours at 35 °C
Controlled environment: 1523813 hours at 35 °C
THERMAL SPECIFICATIONS
Table 5 provides the thermal specifications that apply to both the Tesla P100 PCIe 16GB
and the Tesla P100 PCIe 12GB.
Table 5. Thermal Specifications
Parameter Value Units
Total board power 250 W
GPU shutdown temperature 85 °C
GPU slowdown temperature (50% clock slowdown) 82 °C
GPU maximum operating temperature 80 °C
Tesla P100 PCIe
GPU Accelerator PB-08248-001_v01 | 4
Specifications
System Airflow Requirements
The Tesla P100 PCIe board employs a bidirectional heat sink, which accepts airflow
either left-to-right or right-to-left directions.
Figure 1. Tesla P100 PCIe Airflow Directions with Optional I/O Bracket
Airflow Rate vs. Heat Sink Inlet Temperature
Table 6 specifies the minimum volumetric airflow rate for the Tesla P100 PCIe boards at
the specified air temperature points measured at the card’s inlet. These data points
represent the conditions under which the board was qualified.
The following table applies to both the Tesla P100-PCIe 16GB and the Tesla P100-PCIe
12GB.
Tesla P100 PCIe
GPU Accelerator PB-08248-001_v01 | 5
Specifications
3
Table 6. Minimum CFM Requirements per Inlet Air Temperature
Heat Sink Inlet Temperature (°C)
Minimum Airflow Rate (CFM)
1,2
Qualified by NVIDIA
30 12
35 15 Yes
40 19 Yes
45 26 Yes
50 37
Any
Notes:
1
The minimum airflow is assumed to be a fully ducted airflow through the product’s heat sink. System
thermal designs must account for any air bypass the system may allow.
2
The airflow and air inlet temperature data provided in this table is a qualification reference point, not an
absolute specification. Refer to the system design guide for more details regarding the system airflow
design guidance.
3
This is the minimum required airflow when input power is provided to the card regardless of heat sink
inlet temperature or the system’s operating mode. For example, if the system is in idle mode but the 12V
input rails to the card are active, the system needs to provide the specified minimum airflow.
12
POWER AND ELECTRICAL SPECIFICATIONS
Table 7 provides the Input EDP Continuous specifications for input rails. The Input EDP
Continuous is specified over a 1 second time scale. The peak input power over a 1
second moving average will not exceed the specified Input EDP Continuous value.
The following table applies to both the Tesla P100-PCIe 16GB and the Tesla P100-PCIe
12GB.
Table 7. Input EDP Continuous Specifications
Input Rails Maximum Value Moving Average
PCI Express edge connector (12V ± 8%)
CPU 8-pin power connector (12V ± 8%)
PCI Express edge connector (3V3 ± 9%)
Total graphic power1
Note:
1
Individual rails are specified at their absolute maximum values, but the total power consumption of
the board during operation will never exceed the specified total graphic power.
66 W 1 s
5.5 A 1 s
240 W 1 s
20 A 1 s
10 W 1 s
3 A 1 s
250 W 1 s
Tesla P100 PCIe
GPU Accelerator PB-08248-001_v01 | 6
DESIGN DISCUSSION
FORM FACTOR
The Tesla P100 PCIe board conforms to NVIDIA Form Factor 3.0 specification. In this
product brief, nominal dimensions are shown; for tolerances, see the attached 2D
mechanical drawings.
Figure 2. Tesla P100 PCIe Board Dimensions with Optional I/O Bracket
Tesla P100 PCIe
GPU Accelerator PB-08248-001_v01 | 7
Design Discussion
POWER CONNECTOR PLACEMENT
The board provides a CPU 8-pin power connector on the East edge of the board.
CPU 8-Pin Power Connector
Figure 3. CPU 8-Pin Power Connector with Optional I/O Bracket
Table 8. Supported Auxiliary Power Connections
Board Connector PSU Cable
CPU 8-pin 1x CPU 8-pin cable
CPU to PCIe 8-pin dongle 2x PCIe 8-pin cable
2x PCIe 6-pin cable
1x PCIe 8-pin cable and 1x PCIe 6-pin cable
Notes:
1
The PCIe 6-pin cable must be capable of carrying up to 120 W.
1
1
Tesla P100 PCIe
GPU Accelerator PB-08248-001_v01 | 8
Design Discussion
CPU 8-Pin to PCIe 8-Pin Dongle
Figure 9 lists the pin assignments of the dongle. The part number for the dongle is:
NVPN: 030-0571-000
Figure 4. CPU 8-Pin to PCIe 8-Pin Dongle
EXTENDERS
The Tesla P100 PCIe board provides two extender options as shown in the following
figures.
NVPN: 682-00003-5555-002 –Long offset extender (Figure 5)
● Card + extender = 339 mm
NVPN: 682-00003-5555-000 – Straight extender (Figure 6)
● Card + extender = 312 mm
Tesla P100 PCIe
GPU Accelerator PB-08248-001_v01 | 9
Figure 5. Long Offset Extender
Design Discussion
Figure 6. Straight Extender
Using the standard NVIDIA extender ensures greatest forward compatibility with
future NVIDIA product offerings.
If the standard extender will not work, OEMs may design a custom attach method
using the extender mounting holes on the heat sink baseplate. The extender mounting
holes will vary among NVIDIA products, so designing for flexibility is
recommended.
Tesla P100 PCIe
GPU Accelerator PB-08248-001_v01 | 10
SUPPORT INFORMATION
CERTIFICATES AND AGENCIES
Certifications
Windows Hardware Quality Lab (WHQL):
● Certified Windows 7, Windows 8.1, and Windows 10
● Certified Windows Server 2008 R2 and Windows Server 2012 R2
Ergonomic requirements for office work W/VDTs (ISO 9241)
EU Reduction of Hazardous Substances (EU RoHS)
Joint Industry guide (J-STD) / Registration, Evaluation, Authorization, and Restriction
of Chemical Substance (EU) – (JIG / REACH)
Halogen Free (HF)
EU Waste Electrical and Electronic Equipment (WEEE)
Agencies
Australian Communications and Media Authority and New Zealand Radio Spectrum
Management (RCM)
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, UL)
Voluntary Control Council for Interference (VCCI)
Tesla P100 PCIe
GPU Accelerator PB-08248-001_v01 | 11
LANGUAGES
Table 9. Languages Supported
Support Information
Languages
English (US) Yes Yes
English (UK) Yes Yes
Arabic Yes
Chinese, Simplified Yes
Chinese, Traditional Yes
Czech Yes
Danish Yes
Dutch Yes
Finnish Yes
French (European) Yes
German Yes
Greek Yes
Hebrew Yes
Hungarian Yes
Italian Yes
Japanese Yes
Korean Yes
Norwegian Yes
Polish Yes
Portuguese (Brazil) Yes
Portuguese (European/Iberian) Yes
Russian Yes
Slovak Yes
Slovenian Yes
Spanish (European) Yes
Spanish (Latin America) Yes
Swedish Yes
Thai Yes
Turkish Yes
Note:
1
Windows 7, Windows 8, Windows 8.1, Windows 10, Windows Server 2008 R2,
and Windows Server 2012 R2 are supported.
Windows1
Linux
Tesla P100 PCIe
GPU Accelerator PB-08248-001_v01 | 12
Notice
The information provided in this specification is believed to be accurate and reliable as of the date provided. However, NVIDIA
Corporation (“NVIDIA”) does not give any representations or warranties, expressed or implied, as to the accuracy or
completeness of such information. NVIDIA shall have no liability for the consequences or use of such information or for any
infringement of patents or other rights of third parties that may result from its use. This publication supersedes and replaces all
other specifications for the product that may have been previously supplied.
NVIDIA reserves the right to make corrections, modifications, enhancements, improvements, and other changes to this
specification, at any time and/or to discontinue any product or service without notice. Customer should obtain the latest
relevant specification before placing orders and should verify that such information is current and complete.
NVIDIA products are sold subject to the NVIDIA standard terms and conditions of sale supplied at the time of order
acknowledgement, unless otherwise agreed in an individual sales agreement signed by authorized representatives of NVIDIA and
customer. NVIDIA hereby expressly objects to applying any customer general terms and conditions with regard to the purchase of
the NVIDIA product referenced in this specification.
NVIDIA products are not designed, authorized or warranted to be suitable for use in medical, military, aircraft, space or life
support equipment, nor in applications where failure or malfunction of the NVIDIA product can reasonably be expected to result
in personal injury, death or property or environmental damage. NVIDIA accepts no liability for inclusion and/or use of NVIDIA
products in such equipment or applications and therefore such inclusion and/or use is at customer’s own risk.
NVIDIA makes no representation or warranty that products based on these specifications will be suitable for any specified use
without further testing or modification. Testing of all parameters of each product is not necessarily performed by NVIDIA. It is
customer’s sole responsibility to ensure the product is suitable and fit for the application planned by customer and to do the
necessary testing for the application in order to avoid a default of the application or the product. Weaknesses in customer’s
product designs may affect the quality and reliability of the NVIDIA product and may result in additional or different conditions
and/or requirements beyond those contained in this specification. NVIDIA does not accept any liability related to any default,
damage, costs or problem which may be based on or attributable to: (i) the use of the NVIDIA product in any manner that is
contrary to this specification, or (ii) customer product designs.
No license, either expressed or implied, is granted under any NVIDIA patent right, copyright, or other NVIDIA intellectual
property right under this specification. Information published by NVIDIA regarding third-party products or services does not
constitute a license from NVIDIA to use such products or services or a warranty or endorsement thereof. Use of such information
may require a license from a third party under the patents or other intellectual property rights of the third party, or a license
from NVIDIA under the patents or other intellectual property rights of NVIDIA. Reproduction of information in this specification is
permissible only if reproduction is approved by NVIDIA in writing, is reproduced without alteration, and is accompanied by all
associated conditions, limitations, and notices.
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. Notwithstanding any damages that customer
might incur for any reason whatsoever, NVIDIA’s aggregate and cumulative liability towards customer for the products described
herein shall be limited in accordance with the NVIDIA terms and conditions of sale for the product.
Trademarks
NVIDIA, the NVIDIA logo, CUDA, NVIDIA GPU Boost, Pascal, and Tesla are trademarks and/or registered trademarks of NVIDIA
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.
Copyright
© 2016 NVIDIA Corporation. All rights reserved.
www.nvidia.com
Loading...