How to request an interactive GPU on ARC - Revision history

Dmitri: /* GPUs on ARC */

2025-04-03T18:18:59Z

GPUs on ARC

← Older revision		Revision as of 18:18, 3 April 2025
Line 1:		Line 1:
	= GPUs on ARC =		= GPUs on ARC =
	Please remember, that interactive jobs on ARC should be limited to '''5 hours or less''.		Please remember, that interactive jobs on ARC should be limited to '''5 hours or less'''.
	Do not request more than 5 hours for your interactive jobs on ARC.		Do not request more than 5 hours for your interactive jobs on ARC.

2025-04-03T18:18:49Z

GPUs on ARC

← Older revision		Revision as of 18:18, 3 April 2025
Line 1:		Line 1:
	= GPUs on ARC =		= GPUs on ARC =
			Please remember, that interactive jobs on ARC should be limited to '''5 hours or less''.
			Do not request more than 5 hours for your interactive jobs on ARC.

	Currently, there are three partitions on ARC that have usable CUDA GPUs, '''gpu-v100''', '''gpu-a100''', and '''bigmem'''.		Currently, there are three partitions on ARC that have usable CUDA GPUs, '''gpu-v100''', '''gpu-a100''', and '''bigmem'''.

2023-09-20T22:45:12Z

Added ARC category

2023-09-20T21:56:55Z

Added categories

2023-03-01T17:25:21Z

← Older revision		Revision as of 17:25, 1 March 2023
Line 1:		Line 1:
			= GPUs on ARC =
	Currently, there are three partitions on ARC that have usable CUDA GPUs, '''gpu-v100''', '''gpu-a100''', and '''bigmem'''.		Currently, there are three partitions on ARC that have usable CUDA GPUs, '''gpu-v100''', '''gpu-a100''', and '''bigmem'''.

2023-03-01T17:24:48Z

← Older revision		Revision as of 17:24, 1 March 2023
Line 19:		Line 19:
	For more information about finding hardware on ARC, see [[How to find available partitions on ARC]].		For more information about finding hardware on ARC, see [[How to find available partitions on ARC]].

			=== Example 1 ===
	1 GPU and 4 CPUs on the '''gpu-v100''' partition for 1 hour (16 GB of RAM):		1 GPU and 4 CPUs on the '''gpu-v100''' partition for 1 hour (16 GB of RAM):
	salloc -N1 -n1 -c4 --mem=16GB --gres=gpu:1 -p gpu-v100 -t 1:00:00		salloc -N1 -n1 -c4 --mem=16GB --gres=gpu:1 -p gpu-v100 -t 1:00:00

			=== Example 2 ===
	1 GPU and 4 CPUs on the '''bigmem''' partition for 1 hour (256 GB of RAM):		1 GPU and 4 CPUs on the '''bigmem''' partition for 1 hour (256 GB of RAM):
	$ salloc -N1 -n1 -c4 --mem=256gb --gres=gpu:1 -p bigmem -t 1:00:00		$ salloc -N1 -n1 -c4 --mem=256gb --gres=gpu:1 -p bigmem -t 1:00:00

			=== Checking the allocated GPU(s) ===
	Use the '''nvidia-smi''' command to check the GPU:		Use the '''nvidia-smi''' command to check the GPU:
	<pre>		<pre>
Line 53:		Line 53:
	</pre>		</pre>

			= Links =
	[[How-Tos]]		[[How-Tos]]

2023-03-01T17:23:24Z

← Older revision		Revision as of 17:23, 1 March 2023
Line 17:		Line 17:
	.........		.........
	</pre>		</pre>
			For more information about finding hardware on ARC, see [[How to find available partitions on ARC]].

2023-03-01T17:22:07Z

← Older revision		Revision as of 17:22, 1 March 2023
Line 1:		Line 1:
	Currently, there are three partitions on ARC that have usable CUDA GPUs, '''gpu-v100''', ~~<nowiki>~~'''~~</nowiki>~~gpu-a100 and '''bigmem'''.		Currently, there are three partitions on ARC that have usable CUDA GPUs, '''gpu-v100''', '''gpu-a100''', and '''bigmem'''.

	The <code>arc.hardware</code> tool can be used to see that.		The <code>arc.hardware</code> tool can be used to see that.

2023-03-01T17:21:43Z

← Older revision		Revision as of 17:21, 1 March 2023
Line 1:		Line 1:
	Currently, there are ~~two~~ partitions on ARC that have usable CUDA GPUs, '''gpu-v100''', and '''bigmem'''.		Currently, there are three partitions on ARC that have usable CUDA GPUs, '''gpu-v100''', <nowiki>'''</nowiki>gpu-a100 and '''bigmem'''.

	The <code>arc.hardware</code> tool can be used to see that.		The <code>arc.hardware</code> tool can be used to see that.

2022-06-03T17:48:18Z