GitHub - AakashHegde/Parallel_Computing: Introductory programs that use the GPU or multiple threads to acheive speedup.

AakashHegde / Parallel_Computing Public

Notifications You must be signed in to change notification settings
Fork 0
Star 0

Introductory programs that use the GPU or multiple threads to acheive speedup.

Notifications

Name		Name	Last commit message	Last commit date
Latest commit History 4 Commits
LICENSE		LICENSE
OpenCL_intro.cpp		OpenCL_intro.cpp
OpenCL_mul.cpp		OpenCL_mul.cpp
OpenMP_BitonicSort.cpp		OpenMP_BitonicSort.cpp
OpenMP_BitonicV2.cpp		OpenMP_BitonicV2.cpp
OpenMP_PiValue.cpp		OpenMP_PiValue.cpp
ReadMe.txt		ReadMe.txt
VectorMul.txt		VectorMul.txt
kernel_hw.txt		kernel_hw.txt

Repository files navigation

The following programs utilise multiple processing units to acheive a speedup.

There are two types of implementations in this project : 
	1) GPGPU programming using OpenCL
	2) Multi-thread programming using OpenMP

OpenCL:
Introductory programs 	- print "hello world"
			- add a vector of large size
Uses an older version of OpenCL API for compatibility reasons.
No speedup is obtained because of how basic the programs are.

OpenMP:
Program to perform Bitonic Sort - a parallel sort algorithm.
Good speedup observed when workload is appropriate.
Compares the speedup for different array sizes, and displays all timings in a tabulated manner.