Streams and Multi-GPU using CUVI - Revision history

Jawad: /* CUVI example */

2022-10-18T07:44:32Z

CUVI example

← Older revision		Revision as of 07:44, 18 October 2022
Line 6:		Line 6:
	{\|		{\|
	\|style="font-size:130%;"\|		\|style="font-size:130%;"\|
	<syntaxhighlight lang="~~cpp~~">		<syntaxhighlight lang="js">

	//You may notice that all the functions below take an optional parameter for stream		//You may notice that all the functions below take an optional parameter for stream

Jawad at 07:37, 17 October 2022

2022-10-17T07:37:00Z

← Older revision		Revision as of 07:37, 17 October 2022
Line 1:		Line 1:
	==Using Streams with CUVI==		==Using Streams with CUVI==
	CUVI framework provides a way to use streams with minimal coding effort. Each function call in CUVI takes an optional parameter to specify the stream on which it should run. The code below shows how a simple function call of CUVI can be divided into streaming calls on GPU. For most of the cases this will result in better performance as copying image data to GPU and processing that data on GPU happens simultaneously		CUVI framework provides a way to use streams with minimal coding effort. Each function call in CUVI takes an optional parameter to specify the stream on which it should run. The code below shows how a simple function call of CUVI can be divided into streaming calls on GPU. For most of the cases this will result in better performance as copying image data to GPU and processing that data on GPU happens simultaneously.

	===CUVI example===		===CUVI example===

Jawad at 07:36, 17 October 2022

2022-10-17T07:36:54Z

← Older revision		Revision as of 07:36, 17 October 2022
Line 1:		Line 1:
	==Using Streams with CUVI==		==Using Streams with CUVI==
	CUVI framework provides a way to use streams with minimal coding effort. Each function call in CUVI takes an optional parameter to specify the stream on which it should run. The code below shows how a simple function call of CUVI can be divided into streaming calls on GPU. For most of the cases this will result in better performance as copying image data to GPU and processing that data on GPU happens simultaneously.		CUVI framework provides a way to use streams with minimal coding effort. Each function call in CUVI takes an optional parameter to specify the stream on which it should run. The code below shows how a simple function call of CUVI can be divided into streaming calls on GPU. For most of the cases this will result in better performance as copying image data to GPU and processing that data on GPU happens simultaneously

	===CUVI example===		===CUVI example===

Ghazanfar: /* CUVI example */

2012-05-04T10:15:08Z

CUVI example

← Older revision		Revision as of 10:15, 4 May 2012
Line 9:		Line 9:

	//You may notice that all the functions below take an optional parameter for stream		//You may notice that all the functions below take an optional parameter for stream
	//If the ~~users~~ doesn't wish to use it the program will execute on a single stream by default		//If the user doesn't wish to use it, the program will execute on a single stream by default

	//Creating a 3-channel Image container on GPU		//Creating a 3-channel Image container on GPU
	CuviImage* gimg = new CuviImage(size,~~img~~->depth,3);		CuviImage* gimg = new CuviImage(size,host_img->depth,3);
	//Creating a ~~sing~~ channel Image container for output on GPU		//Creating a single channel Image container for output on GPU
	CuviImage* gout = new CuviImage(size,~~img~~->depth,1);		CuviImage* gout = new CuviImage(size,host_out->depth,1);

	//Uploading RGB image to GPU		//Uploading RGB image to GPU
Line 27:		Line 27:
	</syntaxhighlight>		</syntaxhighlight>
	\|}		\|}


	===Same example with Streams===		===Same example with Streams===

Jawad: /* Multi-GPU in CUVI */

2012-05-04T10:08:58Z

Multi-GPU in CUVI

← Older revision		Revision as of 10:08, 4 May 2012
Line 92:		Line 92:
	\|}		\|}

	A user can divide stream execution among GPUs just by selecting the desired device for execution before any CUVI call. One does that by calling <code>cuviSetCurrentDevice(X);</code> where 'X' represents the device id and its range is {0,N-1} in a machine containing N CUDA capable GPUs. Any ~~function~~ call of CUVI following this will execute on device X until you set the current device to another.		A user can divide stream execution among GPUs just by selecting the desired device for execution before any CUVI call. One does that by calling <code>cuviSetCurrentDevice(X);</code> where 'X' represents the device id and its range is {0,N-1} in a machine containing N CUDA capable GPUs. Any call of CUVI following this will execute on device X until you set the current device to another.

Jawad: /* Multi-GPU in CUVI */

2012-05-04T10:08:11Z

Multi-GPU in CUVI

← Older revision		Revision as of 10:08, 4 May 2012
Line 92:		Line 92:
	\|}		\|}

	~~You~~ can divide stream execution among GPUs just by selecting the device for execution ~~using~~ <code>cuviSetCurrentDevice(X);</code> ~~before any CUVI call.~~ 'X' represents the device id and its range is {0,N-1} in a machine containing N CUDA capable GPUs. Any function call of CUVI following ~~<code>cuviSetCurrentDevice(X)</code>~~ will execute on device X until you set the current device to another.		A user can divide stream execution among GPUs just by selecting the desired device for execution before any CUVI call. One does that by calling <code>cuviSetCurrentDevice(X);</code> where 'X' represents the device id and its range is {0,N-1} in a machine containing N CUDA capable GPUs. Any function call of CUVI following this will execute on device X until you set the current device to another.

Jawad: /* Multi-GPU in CUVI */

2012-05-04T10:05:43Z

Multi-GPU in CUVI

← Older revision		Revision as of 10:05, 4 May 2012
Line 81:		Line 81:

	==Multi-GPU in CUVI==		==Multi-GPU in CUVI==
	Applications that use CUVI also have the liberty to scale up on a Multi-GPU environment without changing a single line of code. ~~We now~~ know how to play with streams in CUVI~~. Multi~~-GPU is nothing more than ~~just~~ dividing those ~~chunk's~~ execution across all the ~~GPU~~ devices ~~installed~~ in the machine. You can write a single piece of code with few checks and error handling that will run of a single GPU machine ~~as well as~~ on ~~Multi~~-GPU ~~machine while using full capability of that machine~~.		Applications that use CUVI also have the liberty to scale up on a Multi-GPU environment without changing a single line of code. Now that we know how to play with streams in CUVI, multi-GPU is nothing more than dividing those stream execution across all the available devices in the machine. You can write a single piece of code with few checks and error handling that will run on a single GPU machine and scale up on a multi-GPU environment.
	{\|		{\|
	\|style="font-size:130%;"\|		\|style="font-size:130%;"\|
Line 92:		Line 92:
	\|}		\|}

	~~If you have more than one GPU installed on the machine you~~ can divide stream execution among ~~them~~ by ~~just~~ selecting the device ~~before any CUVI call~~ using <code>cuviSetCurrentDevice(X);</code> ~~where~~ X is the device id and its range is {0,N-1} in a machine containing N CUDA capable GPUs. Any function call of CUVI following <code>cuviSetCurrentDevice(X)</code> will execute on ~~the~~ device X until you set the current device to another.		You can divide stream execution among GPUs just by selecting the device for execution using <code>cuviSetCurrentDevice(X);</code> before any CUVI call. 'X' represents the device id and its range is {0,N-1} in a machine containing N CUDA capable GPUs. Any function call of CUVI following <code>cuviSetCurrentDevice(X)</code> will execute on device X until you set the current device to another.

Jawad at 10:01, 4 May 2012

2012-05-04T10:01:03Z

← Older revision		Revision as of 10:01, 4 May 2012
Line 91:		Line 91:
	</syntaxhighlight>		</syntaxhighlight>
	\|}		\|}

			If you have more than one GPU installed on the machine you can divide stream execution among them by just selecting the device before any CUVI call using <code>cuviSetCurrentDevice(X);</code> where X is the device id and its range is {0,N-1} in a machine containing N CUDA capable GPUs. Any function call of CUVI following <code>cuviSetCurrentDevice(X)</code> will execute on the device X until you set the current device to another.

Jawad at 09:56, 4 May 2012

2012-05-04T09:56:02Z

← Older revision		Revision as of 09:56, 4 May 2012
Line 81:		Line 81:

	==Multi-GPU in CUVI==		==Multi-GPU in CUVI==
			Applications that use CUVI also have the liberty to scale up on a Multi-GPU environment without changing a single line of code. We now know how to play with streams in CUVI. Multi-GPU is nothing more than just dividing those chunk's execution across all the GPU devices installed in the machine. You can write a single piece of code with few checks and error handling that will run of a single GPU machine as well as on Multi-GPU machine while using full capability of that machine.
			{\|
			\|style="font-size:130%;"\|
			<syntaxhighlight lang="cpp">
			cuviGetDeviceCount(&DeviceCount);
			if(DeviceCount>1){
			//Multi-GPU code here
			}
			</syntaxhighlight>
			\|}

Jawad at 09:49, 4 May 2012

2012-05-04T09:49:37Z

← Older revision		Revision as of 09:49, 4 May 2012
Line 78:		Line 78:
	</syntaxhighlight>		</syntaxhighlight>
	\|}		\|}


			==Multi-GPU in CUVI==