CUDA - Revision history

Nathanrwells: /* CUDA Overview */

2025-06-26T13:44:37Z

CUDA Overview

← Older revision		Revision as of 08:44, 26 June 2025
Line 1:		Line 1:
	== CUDA Overview ==		== CUDA Overview ==
	[[wikipedia:CUDA\|CUDA]] is a feature set for programming nVidia [[wikipedia:Graphics_processing_unit\|GPUs]]. We have many dwarf nodes that are CUDA-enabled with 1-2 GPUs and most of the Wizard nodes have 4 GPUs each. Most of these are consumer grade [https://www.nvidia.com/en-us/geforce/products/10series/geforce-gtx-1080-ti/ nVidia 1080 Ti graphics cards] that are good for accelerating 32-bit calculations. Dwarf36-38 have two [https://www.nvidia.com/en-us/design-visualization/rtx-a4000/ nVidia RTX A4000 graphic cards] and dwarf39 has two [https://www.nvidia.com/en-us/geforce/products/10series/geforce-gtx-1080-ti/ nVidia 1080 Ti graphics cards] that are available for anybody to use but you'll need to ~~email beocathelp@ksu.edu or~~ contact Beocat staff through a [https://support.ksu.edu/TDClient/30/Portal/Requests/ServiceDet?ID=44 TDX Ticket] to request being added to the GPU priority group then you'll need to submit jobs with <B>--partition=ksu-gen-gpu.q</B>. Wizard20 and wizard21 each have two [https://www.nvidia.com/object/quadro-graphics-with-pascal.html nVidia P100 cards] that are much more costly than the consumer grade 1080Ti cards but can accelerate 64-bit calculations much better.		[[wikipedia:CUDA\|CUDA]] is a feature set for programming nVidia [[wikipedia:Graphics_processing_unit\|GPUs]]. We have many dwarf nodes that are CUDA-enabled with 1-2 GPUs and most of the Wizard nodes have 4 GPUs each. Most of these are consumer grade [https://www.nvidia.com/en-us/geforce/products/10series/geforce-gtx-1080-ti/ nVidia 1080 Ti graphics cards] that are good for accelerating 32-bit calculations. Dwarf36-38 have two [https://www.nvidia.com/en-us/design-visualization/rtx-a4000/ nVidia RTX A4000 graphic cards] and dwarf39 has two [https://www.nvidia.com/en-us/geforce/products/10series/geforce-gtx-1080-ti/ nVidia 1080 Ti graphics cards] that are available for anybody to use but you'll need to contact Beocat staff through a [https://support.ksu.edu/TDClient/30/Portal/Requests/ServiceDet?ID=44 TDX Ticket] or email beocat@cs.ksu.edu to request being added to the GPU priority group then you'll need to submit jobs with <B>--partition=ksu-gen-gpu.q</B>. Wizard20 and wizard21 each have two [https://www.nvidia.com/object/quadro-graphics-with-pascal.html nVidia P100 cards] that are much more costly than the consumer grade 1080Ti cards but can accelerate 64-bit calculations much better.

	== Training videos ==		== Training videos ==

Nathanrwells at 20:16, 23 June 2025

2025-06-23T20:16:35Z

← Older revision		Revision as of 15:16, 23 June 2025
Line 1:		Line 1:
	== CUDA Overview ==		== CUDA Overview ==
	[[wikipedia:CUDA\|CUDA]] is a feature set for programming nVidia [[wikipedia:Graphics_processing_unit\|GPUs]]. We have many dwarf nodes that are CUDA-enabled with 1-2 GPUs and most of the Wizard nodes have 4 GPUs each. Most of these are consumer grade [https://www.nvidia.com/en-us/geforce/products/10series/geforce-gtx-1080-ti/ nVidia 1080 Ti graphics cards] that are good for accelerating 32-bit calculations. Dwarf36-38 have two [https://www.nvidia.com/en-us/design-visualization/rtx-a4000/ nVidia RTX A4000 graphic cards] and dwarf39 has two [https://www.nvidia.com/en-us/geforce/products/10series/geforce-gtx-1080-ti/ nVidia 1080 Ti graphics cards] that are available for anybody to use but you'll need to email ~~beocat~~@cs.ksu.edu to request being added to the GPU priority group then you'll need to submit jobs with <B>--partition=ksu-gen-gpu.q</B>. Wizard20 and wizard21 each have two [https://www.nvidia.com/object/quadro-graphics-with-pascal.html nVidia P100 cards] that are much more costly than the consumer grade 1080Ti cards but can accelerate 64-bit calculations much better.		[[wikipedia:CUDA\|CUDA]] is a feature set for programming nVidia [[wikipedia:Graphics_processing_unit\|GPUs]]. We have many dwarf nodes that are CUDA-enabled with 1-2 GPUs and most of the Wizard nodes have 4 GPUs each. Most of these are consumer grade [https://www.nvidia.com/en-us/geforce/products/10series/geforce-gtx-1080-ti/ nVidia 1080 Ti graphics cards] that are good for accelerating 32-bit calculations. Dwarf36-38 have two [https://www.nvidia.com/en-us/design-visualization/rtx-a4000/ nVidia RTX A4000 graphic cards] and dwarf39 has two [https://www.nvidia.com/en-us/geforce/products/10series/geforce-gtx-1080-ti/ nVidia 1080 Ti graphics cards] that are available for anybody to use but you'll need to email beocathelp@ksu.edu or contact Beocat staff through a [https://support.ksu.edu/TDClient/30/Portal/Requests/ServiceDet?ID=44 TDX Ticket] to request being added to the GPU priority group then you'll need to submit jobs with <B>--partition=ksu-gen-gpu.q</B>. Wizard20 and wizard21 each have two [https://www.nvidia.com/object/quadro-graphics-with-pascal.html nVidia P100 cards] that are much more costly than the consumer grade 1080Ti cards but can accelerate 64-bit calculations much better.

	== Training videos ==		== Training videos ==

Mozes at 00:41, 9 December 2022

2022-12-09T00:41:48Z

← Older revision		Revision as of 19:41, 8 December 2022
Line 4:		Line 4:
	== Training videos ==		== Training videos ==
	CUDA Programming Model Overview: [http://www.youtube.com/watch?v=aveYOlBSe-Y http://www.youtube.com/watch?v=aveYOlBSe-Y]		CUDA Programming Model Overview: [http://www.youtube.com/watch?v=aveYOlBSe-Y http://www.youtube.com/watch?v=aveYOlBSe-Y]

	{{#widget:YouTube\|id=aveYOlBSe-Y\|width=800\|height=600}}		{{#widget:YouTube\|id=aveYOlBSe-Y\|width=800\|height=600}}

	CUDA Programming Basics Part I (Host functions): [http://www.youtube.com/watch?v=79VARRFwQgY http://www.youtube.com/watch?v=79VARRFwQgY]		CUDA Programming Basics Part I (Host functions): [http://www.youtube.com/watch?v=79VARRFwQgY http://www.youtube.com/watch?v=79VARRFwQgY]

	{{#widget:YouTube\|id=79VARRFwQgY\|width=800\|height=600}}		{{#widget:YouTube\|id=79VARRFwQgY\|width=800\|height=600}}

	CUDA Programming Basics Part II (Device functions): [http://www.youtube.com/watch?v=G5-iI1ogDW4 http://www.youtube.com/watch?v=G5-iI1ogDW4]		CUDA Programming Basics Part II (Device functions): [http://www.youtube.com/watch?v=G5-iI1ogDW4 http://www.youtube.com/watch?v=G5-iI1ogDW4]

	{{#widget:YouTube\|id=G5-iI1ogDW4\|width=800\|height=600}}		{{#widget:YouTube\|id=G5-iI1ogDW4\|width=800\|height=600}}
	== Compiling CUDA Applications ==		== Compiling CUDA Applications ==

Mozes at 00:41, 9 December 2022

2022-12-09T00:41:25Z

← Older revision		Revision as of 19:41, 8 December 2022
Line 4:		Line 4:
	== Training videos ==		== Training videos ==
	CUDA Programming Model Overview: [http://www.youtube.com/watch?v=aveYOlBSe-Y http://www.youtube.com/watch?v=aveYOlBSe-Y]		CUDA Programming Model Overview: [http://www.youtube.com/watch?v=aveYOlBSe-Y http://www.youtube.com/watch?v=aveYOlBSe-Y]
	~~<HTML5video type~~=~~"youtube"~~ width="800" height=~~"480" autoplay="false">aveYOlBSe-Y</HTML5video>~~		{{#widget:YouTube\|id=aveYOlBSe-Y\|width=800\|height=600}}

	CUDA Programming Basics Part I (Host functions): [http://www.youtube.com/watch?v=79VARRFwQgY http://www.youtube.com/watch?v=79VARRFwQgY]		CUDA Programming Basics Part I (Host functions): [http://www.youtube.com/watch?v=79VARRFwQgY http://www.youtube.com/watch?v=79VARRFwQgY]
	~~<HTML5video type~~=~~"youtube"~~ width="800" height=~~"480" autoplay="false">79VARRFwQgY</HTML5video>~~		{{#widget:YouTube\|id=79VARRFwQgY\|width=800\|height=600}}

	CUDA Programming Basics Part II (Device functions): [http://www.youtube.com/watch?v=G5-iI1ogDW4 http://www.youtube.com/watch?v=G5-iI1ogDW4]		CUDA Programming Basics Part II (Device functions): [http://www.youtube.com/watch?v=G5-iI1ogDW4 http://www.youtube.com/watch?v=G5-iI1ogDW4]
	~~<HTML5video type~~=~~"youtube"~~ width="800" height=~~"480" autoplay="false">G5-iI1ogDW4</HTML5video>~~		{{#widget:YouTube\|id=G5-iI1ogDW4\|width=800\|height=600}}
	== Compiling CUDA Applications ==		== Compiling CUDA Applications ==
	nvcc is the compiler for CUDA applications. When compiling your applications manually you will need to load a CUDA enabled compiler toolchain (e.g. fosscuda):		nvcc is the compiler for CUDA applications. When compiling your applications manually you will need to load a CUDA enabled compiler toolchain (e.g. fosscuda):

Mozes at 00:25, 9 December 2022

2022-12-09T00:25:16Z

← Older revision		Revision as of 19:25, 8 December 2022
Line 1:		Line 1:
	== CUDA Overview ==		== CUDA Overview ==
	[[wikipedia:CUDA\|CUDA]] is a feature set for programming nVidia [[wikipedia:Graphics_processing_unit\|GPUs]]. We have many dwarf nodes that are CUDA-enabled with 1-2 GPUs and most of the Wizard nodes have 4 GPUs each. Most of these are consumer grade [https://www.nvidia.com/en-us/geforce/products/10series/geforce-gtx-1080-ti/ nVidia 1080 Ti graphics cards] that are good for accelerating 32-bit calculations. ~~Dwarf38 has~~ two [https://www.~~geforce~~.com/~~hardware~~/~~desktop~~-~~gpus~~/~~geforce-gtx-980~~-ti/~~specifications~~ nVidia ~~980 Ti~~ graphic cards] and dwarf39 has two [https://www.nvidia.com/en-us/geforce/products/10series/geforce-gtx-1080-ti/ nVidia 1080 Ti graphics cards] that are available for anybody to use but you'll need to email beocat@cs.ksu.edu to request being added to the GPU priority group then you'll need to submit jobs with <B>--partition=ksu-gen-gpu.q</B>. Wizard20 and wizard21 each have two [https://www.nvidia.com/object/quadro-graphics-with-pascal.html nVidia P100 cards] that are much more costly than the consumer grade 1080Ti cards but can accelerate 64-bit calculations much better.		[[wikipedia:CUDA\|CUDA]] is a feature set for programming nVidia [[wikipedia:Graphics_processing_unit\|GPUs]]. We have many dwarf nodes that are CUDA-enabled with 1-2 GPUs and most of the Wizard nodes have 4 GPUs each. Most of these are consumer grade [https://www.nvidia.com/en-us/geforce/products/10series/geforce-gtx-1080-ti/ nVidia 1080 Ti graphics cards] that are good for accelerating 32-bit calculations. Dwarf36-38 have two [https://www.nvidia.com/en-us/design-visualization/rtx-a4000/ nVidia RTX A4000 graphic cards] and dwarf39 has two [https://www.nvidia.com/en-us/geforce/products/10series/geforce-gtx-1080-ti/ nVidia 1080 Ti graphics cards] that are available for anybody to use but you'll need to email beocat@cs.ksu.edu to request being added to the GPU priority group then you'll need to submit jobs with <B>--partition=ksu-gen-gpu.q</B>. Wizard20 and wizard21 each have two [https://www.nvidia.com/object/quadro-graphics-with-pascal.html nVidia P100 cards] that are much more costly than the consumer grade 1080Ti cards but can accelerate 64-bit calculations much better.

	== Training videos ==		== Training videos ==

Mozes: /* Run Your Application */

2020-03-20T17:11:24Z

Run Your Application

← Older revision		Revision as of 12:11, 20 March 2020
Line 82:		Line 82:
	</syntaxhighlight>		</syntaxhighlight>

	Assuming you don't want to run the program interactively because this is a large job, you can submit a job via sbatch, just be sure to add ~~the~~ '<tt>--gres=gpu:1</tt>' directive.		Assuming you don't want to run the program interactively because this is a large job, you can submit a job via sbatch, just be sure to add '<tt>--gres=gpu:1</tt>' to the '''sbatch''' directive.

Mozes: /* Compile Your Application */

2020-03-20T17:09:23Z

Compile Your Application

← Older revision		Revision as of 12:09, 20 March 2020
Line 71:		Line 71:
	=== Compile Your Application ===		=== Compile Your Application ===
	<syntaxhighlight lang="bash">		<syntaxhighlight lang="bash">
	nvcc ~~-I /opt/cuda/sdk/common/inc -arch sm_30~~ vecadd.cu -o vecadd		module load fosscuda
			nvcc vecadd.cu -o vecadd
	</syntaxhighlight>		</syntaxhighlight>
	This will create a program with the name 'vecadd' (specified by the '-o' flag).		This will create a program with the name 'vecadd' (specified by the '-o' flag).

	=== Run Your Application ===		=== Run Your Application ===
	Run the program as you usually would, namely		Run the program as you usually would, namely

Mozes: /* Compiling CUDA Applications */

2020-03-20T17:08:55Z

Compiling CUDA Applications

← Older revision		Revision as of 12:08, 20 March 2020
Line 12:		Line 12:
	<HTML5video type="youtube" width="800" height="480" autoplay="false">G5-iI1ogDW4</HTML5video>		<HTML5video type="youtube" width="800" height="480" autoplay="false">G5-iI1ogDW4</HTML5video>
	== Compiling CUDA Applications ==		== Compiling CUDA Applications ==
	nvcc is the compiler for CUDA applications. When compiling your applications manually you will need to ~~keep 3 things in mind~~:		nvcc is the compiler for CUDA applications. When compiling your applications manually you will need to load a CUDA enabled compiler toolchain (e.g. fosscuda):

	* ~~The CUDA development headers are located here: /opt/cuda/sdk/common/inc~~		* module load fosscuda
	* The CUDA architecture is: sm_30
	* The CUDA SDK is currently not available on the headnode. (compile on the nodes with CUDA, either in your jobscript or interactively via <B>srun -n 1 --gres=gpu:1 --pty bash</B>)
	* '''Do not run your cuda applications on the headnode. I cannot guarantee it will run, and it will give you terrible results if it does run.'''		* '''Do not run your cuda applications on the headnode. I cannot guarantee it will run, and it will give you terrible results if it does run.'''

	~~Putting it all together~~ you can compile CUDA applications as follows:		With those two things in mind, you can compile CUDA applications as follows:

	<syntaxhighlight lang="bash">		<syntaxhighlight lang="bash">
	nvcc ~~-I /opt/cuda/sdk/common/inc -arch sm_30~~ <source>.cu -o <output>		module load fosscuda
			nvcc <source>.cu -o <output>
	</syntaxhighlight>		</syntaxhighlight>

Mozes: /* Run Your Application */

2019-05-10T16:12:34Z

Run Your Application

← Older revision		Revision as of 11:12, 10 May 2019
Line 81:		Line 81:
	</syntaxhighlight>		</syntaxhighlight>

	Assuming you don't want to run the program interactively because this is a large job, you can submit a job via ~~qsub~~, just be sure to add the '<tt>-~~l cuda~~=~~true~~</tt>' directive.		Assuming you don't want to run the program interactively because this is a large job, you can submit a job via sbatch, just be sure to add the '<tt>--gres=gpu:1</tt>' directive.

Daveturner: /* Compiling CUDA Applications */

2019-05-07T22:16:40Z

Compiling CUDA Applications

← Older revision		Revision as of 17:16, 7 May 2019
Line 16:		Line 16:
	* The CUDA development headers are located here: /opt/cuda/sdk/common/inc		* The CUDA development headers are located here: /opt/cuda/sdk/common/inc
	* The CUDA architecture is: sm_30		* The CUDA architecture is: sm_30
	* The CUDA SDK is currently not available on the headnode. (compile on the nodes with CUDA, either in your jobscript or interactively via <tt>srun ~~-N 1~~ -n 1 --gres=gpu:1 --pty bash</tt>)		* The CUDA SDK is currently not available on the headnode. (compile on the nodes with CUDA, either in your jobscript or interactively via <B>srun -n 1 --gres=gpu:1 --pty bash</B>)
	* '''Do not run your cuda applications on the headnode. I cannot guarantee it will run, and it will give you terrible results if it does run.'''		* '''Do not run your cuda applications on the headnode. I cannot guarantee it will run, and it will give you terrible results if it does run.'''