Table 6. Mean Time Between Failure (MTBF) .............................................. 13
Table 7. Languages Supported ............................................................... 15
NVIDIA GRID K2
Graphics Board BD-06580-001_v02 | iv
Page 5
OVERVIEW
The NVIDIA GRID™ K2 is a dual-slot 10.5 inch PCI Express Gen3 graphics card with two
®
high-end NVIDIA
has 8 GB of GDDR5 memory (4 GB per GPU), and a 225 W maximum power limit. The
NVIDIA GRID K2 graphics board uses a passive heat sink that requires system airflow
to properly operate the card within thermal limits. It is designed to accelerate graphics
in virtual remote workstation and virtual desktop environments.
Kepler™ graphics processing units (GPUs). The NVIDIA GRID K2
The NVIDIA GRID K2 graphics board can be configured to enable or disable ECC (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%. ECC protection is for DRAM only.
One 6-pin PCI Express power connector
One 8-pin PCI Express power connector
Overview
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
2 MBit Serial ROM
Virtualization Solutions
Citrix XenServer + XenDesktop with HDX 3D Pro
Citrix XenServer with NVIDIA NVIDIA GRID Hypervisor + XenDesktop with HDX
Microsoft Windows Server 2012 + RemoteFX
Microsoft Windows Server 2008 R2 + RemoteFX
VMware ESXi + View with vSGA
The NVIDIA GRID K2 board (Figure 2) conforms to the PCI Express Gen3 ×16 (4.376
inches by 10.5 inches) form factor. Figure 2 shows the board without the cover.
4.376 inches
Figure 2. NVIDIA GRID K2 Board without Cover
10.5 inches
Note: NVIDIA GRID K2 production boards will ship with a cover.
The NVIDIA GRID K2 board features a vented bracket, as shown in Figure 3. If you are
an OEM who qualifies for bracket modifications, you have the option of receiving your
modules with no bracket installed.
The NVIDIA GRID K2 board utilizes power from both the PCI Express connector and
the auxiliary power connectors. The NVIDIA GRID K2 board supports the following
internal connectors:
One 6-pin PCI Express power connector
One 8-pin PCI Express power connector
Figure 6 and Figure 7 shows the specifications and Table 2 and Table 3 show the pinouts
for the 6-pin and 8-pin PCI Express power connectors, respectively.
The NVIDIA GRID K2 board utilizes power from the PCI Express connector as well as
one or two auxiliary power connectors. Table 4 lists the supported configurations
Table 4. Configurations 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 150 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: If the auxiliary power cables are connected in an unsupported configuration, the NVIDIA GRID K2
board will power up in a low performance mode. The software will de t e ct and re p ort the incorrect
power connections so actions can be taken by th e user to resolve.
Note: The power breakdown per input rail is available to authorized system
partners in the NVIDIA GRID K2 System Design Guide (DG-06546-001).
The NVIDIA GRID K2 graphics board uses passive heat sinks that require system
airflow to properly operate the card within thermal limits. Table 5 provides thermal
information necessary to deliver reliable operation of the NVIDIA GRID K2 GPU. This
information is not intended to provide a specific thermal management solution.
For more detailed information regarding thermal specifications for the NVIDIA GRID
K2 board, refer to the NVIDIA GRID K2 System Design Guide (DG-06546-001).
Table 5. Thermal Specifications
Parameter Value Units
TDP reference operating points1
• Air inlet temperature (at TDP)
• Minimum airflow through opening (full ducted)
2
GPU maximum TDP operating temperature3
2
45
20
92 °C
GPU slowdown temperature (maximum Tj) 92 °C
GPU shutdown temperature 97 °C
Notes:
1
This airflow information is provided as gu idance and is valid only f o r the conditions describe d in the
NVIDIA GRID K2 System Design Guide (DG-06546-001).
2
The airflow and air inl e t temperature data provided here are ref erence points, not ab sol u te
specifications. Refer to the NVIDIA GRID K2 System Design Guide (DG-06546-001) fo r m ore details
regarding the sy stem airflow design g uidance.
3
The GPU maximum TDP operating t e m p erature is the maximu m G PU temperature at which the card is
guaranteed to operate at the total board power level.
The mean time between failure (MTBF) ratings for the NVIDIA GRID K2 are tabulated
in Table 6. The calculation of these values uses the Bellcore’s Parts Count method in
controlled environments.
Table 6. Mean Time Between Failure (MTBF)
Condition (Bellcore Code) MTBF
Ground Benign (GB) environment, 35 °C * 232,528 hours
Ground Fixed (GF) environment, 35 °C ** 127,664 hours
Notes:
*Bellcore Code GB relates to non-mobile equ ipment used in ideal en v ir on m e n t (lab, medical, an d t est
equipment).
**Bellcore Code GF relates to non-mobile equipment used in less t h an ideal environment s ( rack mou n t or
other instrumentation or equipment used in buildings without controlled temperatures).
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)
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® software is only supported in English (U.S. )
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 approval of N VIDIA Corporation.
Trademarks
NVIDIA, the NVIDIA logo, CUDA, Kepler, and NVIDIA GRID 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.