James at 20:11, 27 October 2022

2022-10-27T20:11:43Z

← Older revision		Revision as of 20:11, 27 October 2022
Line 63:		Line 63:
	</pre>		</pre>

	Once compiled, the 'vectorcopy.exe' executable can be run using the following ~~PBS~~ script:		Once compiled, the 'vectorcopy.exe' executable can be run using the following SLURM script:

	<pre>		<pre>
	#!/bin/bash		#!/bin/bash
	#~~PBS~~ -q production_gpu		#SLURM -q production_gpu
	#~~PBS~~ -N THRUST_vcopy		#SLURM -N THRUST_vcopy
	#~~PBS~~ -l select=1:ncpus=1:ngpus=1		#SLURM -l select=1:ncpus=1:ngpus=1
	#~~PBS~~ -l place=free		#SLURM -l place=free
	#~~PBS~~ -V		#SLURM -V

	# Find out which compute node the job is using		# Find out which compute node the job is using
Line 79:		Line 79:

	echo ""		echo ""
	echo "~~PBS~~ node file is located here ... " $~~PBS_NODEFILE~~		echo "SLURM node file is located here ... " $SLURM_NODEFILE
	echo -n "~~PBS~~ node file contains ... "		echo -n "SLURM node file contains ... "
	cat $~~PBS_NODEFILE~~		cat $SLURM_NODEFILE
	echo ""		echo ""

	# Change to working directory		# Change to working directory
	cd $~~PBS_O_WORKDIR~~		cd $SLURM_O_WORKDIR

	# Running executable on a single, gpu-enabled		# Running executable on a single, gpu-enabled

James: Created page with "Thrust provides a rich collection of data parallel primitives such as scan, sort, and reduce, which can be combined together to implement complex algorithms with concise, readable source code. By describing your computation in terms of these high-level abstractions you provide Thrust with the freedom to select the most efficient implementation automatically. As a result, Thrust can be utilized in rapid prototyping of CUDA applications, where programmer productivity matte..."

2022-10-20T19:02:00Z

Created page with "Thrust provides a rich collection of data parallel primitives such as scan, sort, and reduce, which can be combined together to implement complex algorithms with concise, readable source code. By describing your computation in terms of these high-level abstractions you provide Thrust with the freedom to select the most efficient implementation automatically. As a result, Thrust can be utilized in rapid prototyping of CUDA applications, where programmer productivity matte..."

New page

Thrust provides a rich collection of data parallel primitives such as scan, sort, and reduce, which can be combined
together to implement complex algorithms with concise, readable source code. By describing your computation in
terms of these high-level abstractions you provide Thrust with the freedom to select the most efficient implementation
automatically. As a result, Thrust can be utilized in rapid prototyping of CUDA applications, where programmer
productivity matters most, as well as in production, where robustness and absolute performance are crucial.

More detail on the Thrust library is available here [http://code.google.com/p/thrust/wiki/QuickStartGuide]. There
are a collection of example codes here [http://code.google.com/p/thrust/downloads/list]. The Thrust Manual is
available here [http://code.google.com/p/thrust/downloads/detail?name=Thrust%20-%20A%20Productivity-Oriented%20Library%20for%20CUDA.pdf]

Here is a basic C++ example code, which creates and fills a vector on the Host, resizes it, copies it to the
Device, modifies it there, and prints out the modified values.

<pre>
#include <thrust/host_vector.h>
#include <thrust/device_vector.h>

#include <iostream>

int main(void)
{
// H has storage for 4 integers
thrust::host_vector<int> H(4);

// initialize individual elements
H[0] = 14;
H[1] = 20;
H[2] = 38;
H[3] = 46;

// H.size() returns the size of vector H
std::cout << "H has size " << H.size() << std::endl;

// print contents of H
for(int i = 0; i < H.size(); i++)
std::cout << "H[" << i << "] = " << H[i] << std::endl;

// resize H
H.resize(2);

std::cout << "H now has size " << H.size() << std::endl;

// Copy host_vector H to device_vector D
thrust::device_vector<int> D = H;

// elements of D can be modified
D[0] = 99;
D[1] = 88;

// print contents of D
for(int i = 0; i < D.size(); i++)
std::cout << "D[" << i << "] = " << D[i] << std::endl;

// H and D are automatically deleted when the function returns
return 0;
}
</pre>

Assuming this source file were called 'vectcopy.cu', it can be compiled on PENZIAS:

<pre>
nvcc -o vectcopy.exe vectcopy.cu
</pre>

Once compiled, the 'vectorcopy.exe' executable can be run using the following PBS script:

<pre>
#!/bin/bash
#PBS -q production_gpu
#PBS -N THRUST_vcopy
#PBS -l select=1:ncpus=1:ngpus=1
#PBS -l place=free
#PBS -V

# Find out which compute node the job is using
echo ""
echo -n "Running job on compute node ... "
hostname

echo ""
echo "PBS node file is located here ... " $PBS_NODEFILE
echo -n "PBS node file contains ... "
cat $PBS_NODEFILE
echo ""

# Change to working directory
cd $PBS_O_WORKDIR

# Running executable on a single, gpu-enabled
# compute node using 1 CPU and 1 GPU.
echo "CUDA job is starting ... "
echo ""

./vectcopy.exe

echo ""
echo "CUDA job is done!"
</pre>

THRUST - Revision history

James at 20:11, 27 October 2022