GPU Hardware - Revision history

Shtsai: /* Connectivity between GPU devices */

2025-10-29T18:28:52Z

Connectivity between GPU devices

← Older revision		Revision as of 14:28, 29 October 2025
Line 592:		Line 592:
	====Connectivity between GPU devices====		====Connectivity between GPU devices====

	Each GPU device on the quad A100 nodes is connected to the other GPU devices on the same node via a 4 ~~NVlinks~~. For example:		Each GPU device on the quad A100 nodes is connected to the other GPU devices on the same node via a 4 NVLinks. For example:

	<pre class="gcomment">		<pre class="gcomment">
Line 619:		Line 619:


	Each GPU device on the quad H100 nodes is connected to the other GPU devices on the same node via a 6 ~~NVlinks~~. For example:		Each GPU device on the quad H100 nodes is connected to the other GPU devices on the same node via a 6 NVLinks. For example:

	<pre class="gcomment">		<pre class="gcomment">
Line 646:		Line 646:


	~~NVlinks~~ are not present on the L4 ~~or other GPU~~ nodes, or between GPU nodes. For example, on a quad L4 node:		NVLinks are not present on the quad L4 nodes, or between GPU nodes. For example, on a quad L4 node:
	<pre class="gcomment">		<pre class="gcomment">
	$ nvidia-smi topo -m		$ nvidia-smi topo -m

Shtsai: /* Connectivity between GPU devices */

2025-10-29T18:27:57Z

Connectivity between GPU devices

← Older revision		Revision as of 14:27, 29 October 2025
Line 592:		Line 592:
	====Connectivity between GPU devices====		====Connectivity between GPU devices====

	Each GPU device on the quad A100 nodes is connected to the other GPU devices on the node via a 4 NVlinks. For example:		Each GPU device on the quad A100 nodes is connected to the other GPU devices on the same node via a 4 NVlinks. For example:

	<pre class="gcomment">		<pre class="gcomment">
Line 619:		Line 619:


	Each GPU device on the quad H100 nodes is connected to the other GPU devices on the node via a 6 NVlinks. For example:		Each GPU device on the quad H100 nodes is connected to the other GPU devices on the same node via a 6 NVlinks. For example:

	<pre class="gcomment">		<pre class="gcomment">

Shtsai at 18:27, 29 October 2025

2025-10-29T18:27:24Z

@@ Line 588: / Line 588: @@
 deviceQuery, CUDA Driver = CUDART, CUDA Driver Version = 12.8, CUDA Runtime Version = 12.8, NumDevs = 1
 Result = PASS
 </pre>

Shtsai at 16:23, 2 August 2025

2025-08-02T16:23:41Z

@@ Line 166: / Line 166: @@
    Compute Mode:
       < Default (multiple host threads can use ::cudaSetDevice() with device simultaneously) >
 </pre>

Shtsai at 16:17, 2 August 2025

2025-08-02T16:17:33Z

← Older revision		Revision as of 12:17, 2 August 2025
Line 14:		Line 14:
	! scope="col" \| Minimum CUDA version supported		! scope="col" \| Minimum CUDA version supported
	! scope="col" \| Double-precision (fp64) support		! scope="col" \| Double-precision (fp64) support
			! scope="col" \| Notes
	\|-		\|-
	\|-		\|-
	\| H100 \|\| 4 \|\| 16896 \|\| 80GB \|\| 9.0 \|\| 11.8 \|\| yes		\| H100 \|\| 4 \|\| 16896 \|\| 80GB \|\| 9.0 \|\| 11.8 \|\| yes \|\| regular and buy-in
	\|-		\|-
	\| L4 \|\| 4 \|\| ~~7424~~ \|\| ~~24GB~~ \|\| 8.9 \|\| 11.8 \|\| effectively no		\| L40S \|\| 4 \|\| 18176 \|\| 44GB \|\| 8.9 \|\| 11.8 \|\| effectively no \|\|buy-in
	\|-		\|-
	\| ~~A100~~ \|\| 4 \|\| ~~6912~~\|\| ~~80GB~~ \|\| 8.0 \|\| 11.0 \|\| ~~yes~~		\| L4 \|\| 4 \|\| 7424 \|\| 24GB \|\| 8.9 \|\| 11.8 \|\| effectively no \|\| regular and buy-in
	\|-		\|-
	\| ~~V100~~ \|\| 1 \|\| ~~5120~~ \|\| ~~32GB~~ \|\| 7.0\|\| 9.0 \|\| yes		\| A100 \|\| 4 \|\| 6912\|\| 80GB \|\| 8.0 \|\| 11.0 \|\| yes \|\| regular and buy-in
	\|-		\|-
	\| V100 \|\| 2 \|\| 5120 \|\| 32GB \|\| 7.0\|\| 9.0 \|\| yes		\| V100 \|\| 1 \|\| 5120 \|\| 32GB \|\| 7.0\|\| 9.0 \|\| yes \|\| buy-in
	\|-		\|-
	\| V100 \|\| 1 \|\| 5120 \|\| ~~16GB~~ \|\| 7.0\|\| 9.0 \|\| yes		\| V100 \|\| 2 \|\| 5120 \|\| 32GB \|\| 7.0\|\| 9.0 \|\| yes \|\| buy-in
	\|-		\|-
	\| ~~V100S~~ \|\| 1 \|\| 5120 \|\| ~~32GB~~ \|\| 7.0 \|\| 9.0 \|\| yes		\| V100 \|\| 1 \|\| 5120 \|\| 16GB \|\| 7.0\|\| 9.0 \|\| yes \|\| buy-in
	\|-		\|-
	\| P100 \|\| 1 \|\| 3584 \|\| 16GB \|\| 6.0 \|\| 8.0 \|\| yes		\| V100S \|\| 1 \|\| 5120 \|\| 32GB \|\| 7.0 \|\| 9.0 \|\| yes \|\| regular and buy-in
			\|-
			\| P100 \|\| 1 \|\| 3584 \|\| 16GB \|\| 6.0 \|\| 8.0 \|\| yes \|\| regular
	\|-		\|-
	\|}		\|}
Line 378:		Line 381:
	</pre>		</pre>
	-->		-->

			*Specifications for the NVIDIA L40S cards (18176 GPU cores per card). The output of NVIDIA SDK deviceQuery for these GPU cards are
			<pre class="gscript">
			Device 0: "NVIDIA L40S"
			CUDA Driver Version / Runtime Version 12.8 / 12.8
			CUDA Capability Major/Minor version number: 8.9
			Total amount of global memory: 45468 MBytes (47677177856 bytes)
			MapSMtoCores for SM 8.9 is undefined. Default to use 128 Cores/SM
			MapSMtoCores for SM 8.9 is undefined. Default to use 128 Cores/SM
			(142) Multiprocessors, (128) CUDA Cores/MP: 18176 CUDA Cores
			GPU Max Clock rate: 2520 MHz (2.52 GHz)
			Memory Clock rate: 9001 Mhz
			Memory Bus Width: 384-bit
			L2 Cache Size: 100663296 bytes
			Maximum Texture Dimension Size (x,y,z) 1D=(131072), 2D=(131072, 65536), 3D=(16384, 16384, 16384)
			Maximum Layered 1D Texture Size, (num) layers 1D=(32768), 2048 layers
			Maximum Layered 2D Texture Size, (num) layers 2D=(32768, 32768), 2048 layers
			Total amount of constant memory: 65536 bytes
			Total amount of shared memory per block: 49152 bytes
			Total number of registers available per block: 65536
			Warp size: 32
			Maximum number of threads per multiprocessor: 1536
			Maximum number of threads per block: 1024
			Max dimension size of a thread block (x,y,z): (1024, 1024, 64)
			Max dimension size of a grid size (x,y,z): (2147483647, 65535, 65535)
			Maximum memory pitch: 2147483647 bytes
			Texture alignment: 512 bytes
			Concurrent copy and kernel execution: Yes with 2 copy engine(s)
			Run time limit on kernels: No
			Integrated GPU sharing Host Memory: No
			Support host page-locked memory mapping: Yes
			Alignment requirement for Surfaces: Yes
			Device has ECC support: Enabled
			Device supports Unified Addressing (UVA): Yes
			Device supports Compute Preemption: Yes
			Supports Cooperative Kernel Launch: Yes
			Supports MultiDevice Co-op Kernel Launch: Yes
			Device PCI Domain ID / Bus ID / location ID: 0 / 73 / 0
			Compute Mode:
			< Default (multiple host threads can use ::cudaSetDevice() with device simultaneously) >
			</pre>

	* Specifications of NVIDIA Tesla (Volta) V100 and V100S GPU cards (5120 GPU cores per card). The output of NVIDIA SDK deviceQuery for these GPU cards are		* Specifications of NVIDIA Tesla (Volta) V100 and V100S GPU cards (5120 GPU cores per card). The output of NVIDIA SDK deviceQuery for these GPU cards are

Shtsai at 17:06, 31 July 2025

2025-07-31T17:06:57Z

← Older revision		Revision as of 13:06, 31 July 2025
Line 41:		Line 41:
	<pre class="gscript">		<pre class="gscript">
	Device 0: "NVIDIA H100 80GB HBM3"		Device 0: "NVIDIA H100 80GB HBM3"
	CUDA Driver Version / Runtime Version 12.7 / 12.7		CUDA Driver Version / Runtime Version 12.8 / 12.8
	CUDA Capability Major/Minor version number: 9.0		CUDA Capability Major/Minor version number: 9.0
	Total amount of global memory: 81008 MBytes (84943110144 bytes)		Total amount of global memory: 81008 MBytes (84943110144 bytes)
Line 84:		Line 84:
	<pre class="gscript">		<pre class="gscript">
	Device 0: "NVIDIA L4"		Device 0: "NVIDIA L4"
	CUDA Driver Version / Runtime Version 12.7 / 12.7		CUDA Driver Version / Runtime Version 12.8 / 12.8
	CUDA Capability Major/Minor version number: 8.9		CUDA Capability Major/Minor version number: 8.9
	Total amount of global memory: 22491 MBytes (23583784960 bytes)		Total amount of global memory: 22491 MBytes (23583784960 bytes)
Line 127:		Line 127:
	<pre class="gscript">		<pre class="gscript">
	Device 0: "NVIDIA A100-SXM-80GB"EPYC 7543		Device 0: "NVIDIA A100-SXM-80GB"EPYC 7543
	CUDA Driver Version / Runtime Version 12.7 / 12.7		CUDA Driver Version / Runtime Version 12.8 / 12.8
	CUDA Capability Major/Minor version number: 8.0		CUDA Capability Major/Minor version number: 8.0
	Total amount of global memory: 81251 MBytes (85198045184 bytes)		Total amount of global memory: 81251 MBytes (85198045184 bytes)
Line 170:		Line 170:
	<pre class="gscript">		<pre class="gscript">
	Device 0: "Tesla P100-PCIE-16GB"		Device 0: "Tesla P100-PCIE-16GB"
	CUDA Driver Version / Runtime Version 12.7 / 12.7		CUDA Driver Version / Runtime Version 12.8 / 12.8
	CUDA Capability Major/Minor version number: 6.0		CUDA Capability Major/Minor version number: 6.0
	Total amount of global memory: 16281 MBytes (17071734784 bytes)		Total amount of global memory: 16281 MBytes (17071734784 bytes)
Line 382:		Line 382:
	<pre class="gscript">		<pre class="gscript">
	Device 0: "Tesla V100-PCIE-16GB"		Device 0: "Tesla V100-PCIE-16GB"
	CUDA Driver Version / Runtime Version 12.7 / 12.7		CUDA Driver Version / Runtime Version 12.8 / 12.8
	CUDA Capability Major/Minor version number: 7.0		CUDA Capability Major/Minor version number: 7.0
	Total amount of global memory: 16160 MBytes (16945512448 bytes)		Total amount of global memory: 16160 MBytes (16945512448 bytes)
Line 417:		Line 417:
	< Default (multiple host threads can use ::cudaSetDevice() with device simultaneously) >		< Default (multiple host threads can use ::cudaSetDevice() with device simultaneously) >

	deviceQuery, CUDA Driver = CUDART, CUDA Driver Version = 11.4, CUDA Runtime Version = 11.1, NumDevs = 1		deviceQuery, CUDA Driver = CUDART, CUDA Driver Version = 12.8, CUDA Runtime Version = 12.8, NumDevs = 1
	Result = PASS		Result = PASS
	</pre>		</pre>
Line 423:		Line 423:
	<pre class="gscript">		<pre class="gscript">
	Device 0: "Tesla V100-PCIE-32GB"		Device 0: "Tesla V100-PCIE-32GB"
	CUDA Driver Version / Runtime Version 12.7 / 12.7		CUDA Driver Version / Runtime Version 12.8 / 12.8
	CUDA Capability Major/Minor version number: 7.0		CUDA Capability Major/Minor version number: 7.0
	Total amount of global memory: 32510 MBytes (34089730048 bytes)		Total amount of global memory: 32510 MBytes (34089730048 bytes)
Line 458:		Line 458:
	< Default (multiple host threads can use ::cudaSetDevice() with device simultaneously) >		< Default (multiple host threads can use ::cudaSetDevice() with device simultaneously) >

	deviceQuery, CUDA Driver = CUDART, CUDA Driver Version = 11.4, CUDA Runtime Version = 11.1, NumDevs = 1		deviceQuery, CUDA Driver = CUDART, CUDA Driver Version = 12.8, CUDA Runtime Version = 12.8, NumDevs = 1
	Result = PASS		Result = PASS
	</pre>		</pre>
Line 464:		Line 464:
	<pre class="gscript">		<pre class="gscript">
	Device 0: "Tesla V100S-PCIE-32GB"		Device 0: "Tesla V100S-PCIE-32GB"
	CUDA Driver Version / Runtime Version 12.7 / 12.7		CUDA Driver Version / Runtime Version 12.8 / 12.8
	CUDA Capability Major/Minor version number: 7.0		CUDA Capability Major/Minor version number: 7.0
	Total amount of global memory: 32510 MBytes (34089730048 bytes)		Total amount of global memory: 32510 MBytes (34089730048 bytes)
Line 499:		Line 499:
	< Default (multiple host threads can use ::cudaSetDevice() with device simultaneously) >		< Default (multiple host threads can use ::cudaSetDevice() with device simultaneously) >

	deviceQuery, CUDA Driver = CUDART, CUDA Driver Version = 12.7, CUDA Runtime Version = 12.7, NumDevs = 1		deviceQuery, CUDA Driver = CUDART, CUDA Driver Version = 12.8, CUDA Runtime Version = 12.8, NumDevs = 1
	Result = PASS		Result = PASS
	</pre>		</pre>

Shtsai at 20:04, 30 January 2025

2025-01-30T20:04:26Z

Shtsai: /* GPU resources on open-access nodes */

2024-10-25T14:37:35Z

GPU resources on open-access nodes

← Older revision		Revision as of 10:37, 25 October 2024
Line 80:		Line 80:


	* 11 nodes with 4 NVIDIA L4 GPU cards each. Each of these nodes has two 64-core AMD EPYC Genoa processors (128 CPU cores per node) and 745GB of RAM. The output of NVIDIA SDK deviceQuery for one such GPU card is		* 12 nodes with 4 NVIDIA L4 GPU cards each. Each of these nodes has two 64-core AMD EPYC Genoa processors (128 CPU cores per node) and 745GB of RAM. The output of NVIDIA SDK deviceQuery for one such GPU card is

	<pre class="gscript">		<pre class="gscript">

Shtsai: /* GPU resources on open-access nodes */

2024-10-25T14:37:19Z

GPU resources on open-access nodes

← Older revision		Revision as of 10:37, 25 October 2024
Line 37:		Line 37:


	* 10 nodes with 4 NVIDIA H100 GPU cards each. Each of these nodes has two 32-core Intel SapphireRapids processors (64 CPU cores per node) and 1TB of RAM. The output of NVIDIA SDK deviceQuery for one such GPU card is		* 12 nodes with 4 NVIDIA H100 GPU cards each. Each of these nodes has two 32-core Intel SapphireRapids processors (64 CPU cores per node) and 1TB of RAM. The output of NVIDIA SDK deviceQuery for one such GPU card is

	<pre class="gscript">		<pre class="gscript">

Shtsai at 14:37, 25 October 2024

2024-10-25T14:37:00Z

← Older revision		Revision as of 10:37, 25 October 2024
Line 37:		Line 37:


	* 10 nodes with 4 NVIDIA H100 GPU cards each. Each of these nodes has two 64-core Intel SapphireRapids processors (~~128~~ CPU cores per node) and 1TB of RAM. The output of NVIDIA SDK deviceQuery for one such GPU card is		* 10 nodes with 4 NVIDIA H100 GPU cards each. Each of these nodes has two 32-core Intel SapphireRapids processors (64 CPU cores per node) and 1TB of RAM. The output of NVIDIA SDK deviceQuery for one such GPU card is

	<pre class="gscript">		<pre class="gscript">

← Older revision		Revision as of 16:04, 30 January 2025
Line 41:		Line 41:
	<pre class="gscript">		<pre class="gscript">
	Device 0: "NVIDIA H100 80GB HBM3"		Device 0: "NVIDIA H100 80GB HBM3"
	CUDA Driver Version / Runtime Version 12.2 / 12.2		CUDA Driver Version / Runtime Version 12.7 / 12.7
	CUDA Capability Major/Minor version number: 9.0		CUDA Capability Major/Minor version number: 9.0
	Total amount of global memory: 81008 MBytes (84943110144 bytes)		Total amount of global memory: 81008 MBytes (84943110144 bytes)
Line 84:		Line 84:
	<pre class="gscript">		<pre class="gscript">
	Device 0: "NVIDIA L4"		Device 0: "NVIDIA L4"
	CUDA Driver Version / Runtime Version 12.2 / 12.2		CUDA Driver Version / Runtime Version 12.7 / 12.7
	CUDA Capability Major/Minor version number: 8.9		CUDA Capability Major/Minor version number: 8.9
	Total amount of global memory: 22491 MBytes (23583784960 bytes)		Total amount of global memory: 22491 MBytes (23583784960 bytes)
Line 127:		Line 127:
	<pre class="gscript">		<pre class="gscript">
	Device 0: "NVIDIA A100-SXM-80GB"EPYC 7543		Device 0: "NVIDIA A100-SXM-80GB"EPYC 7543
	CUDA Driver Version / Runtime Version 12.2 / 12.2		CUDA Driver Version / Runtime Version 12.7 / 12.7
	CUDA Capability Major/Minor version number: 8.0		CUDA Capability Major/Minor version number: 8.0
	Total amount of global memory: 81251 MBytes (85198045184 bytes)		Total amount of global memory: 81251 MBytes (85198045184 bytes)
Line 170:		Line 170:
	<pre class="gscript">		<pre class="gscript">
	Device 0: "Tesla P100-PCIE-16GB"		Device 0: "Tesla P100-PCIE-16GB"
	CUDA Driver Version / Runtime Version 12.2 / 12.2		CUDA Driver Version / Runtime Version 12.7 / 12.7
	CUDA Capability Major/Minor version number: 6.0		CUDA Capability Major/Minor version number: 6.0
	Total amount of global memory: 16281 MBytes (17071734784 bytes)		Total amount of global memory: 16281 MBytes (17071734784 bytes)
Line 382:		Line 382:
	<pre class="gscript">		<pre class="gscript">
	Device 0: "Tesla V100-PCIE-16GB"		Device 0: "Tesla V100-PCIE-16GB"
	CUDA Driver Version / Runtime Version 12.2 / 12.2		CUDA Driver Version / Runtime Version 12.7 / 12.7
	CUDA Capability Major/Minor version number: 7.0		CUDA Capability Major/Minor version number: 7.0
	Total amount of global memory: 16160 MBytes (16945512448 bytes)		Total amount of global memory: 16160 MBytes (16945512448 bytes)
Line 423:		Line 423:
	<pre class="gscript">		<pre class="gscript">
	Device 0: "Tesla V100-PCIE-32GB"		Device 0: "Tesla V100-PCIE-32GB"
	CUDA Driver Version / Runtime Version 12.2 / 12.2		CUDA Driver Version / Runtime Version 12.7 / 12.7
	CUDA Capability Major/Minor version number: 7.0		CUDA Capability Major/Minor version number: 7.0
	Total amount of global memory: 32510 MBytes (34089730048 bytes)		Total amount of global memory: 32510 MBytes (34089730048 bytes)
Line 464:		Line 464:
	<pre class="gscript">		<pre class="gscript">
	Device 0: "Tesla V100S-PCIE-32GB"		Device 0: "Tesla V100S-PCIE-32GB"
	CUDA Driver Version / Runtime Version 12.2 / 12.2		CUDA Driver Version / Runtime Version 12.7 / 12.7
	CUDA Capability Major/Minor version number: 7.0		CUDA Capability Major/Minor version number: 7.0
	Total amount of global memory: 32510 MBytes (34089730048 bytes)		Total amount of global memory: 32510 MBytes (34089730048 bytes)
Line 499:		Line 499:
	< Default (multiple host threads can use ::cudaSetDevice() with device simultaneously) >		< Default (multiple host threads can use ::cudaSetDevice() with device simultaneously) >

	deviceQuery, CUDA Driver = CUDART, CUDA Driver Version = 11.4, CUDA Runtime Version = 11.1, NumDevs = 1		deviceQuery, CUDA Driver = CUDART, CUDA Driver Version = 12.7, CUDA Runtime Version = 12.7, NumDevs = 1
	Result = PASS		Result = PASS
	</pre>		</pre>