Skip to content
/ xgemm Public

A collection of GEMM benchmarks to verify Summit V100 and Frontier MI-100/MI-200 GPUs.

Notifications You must be signed in to change notification settings

fwang2/xgemm

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

12 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

xGEMM

A collection of GEMM benchmarks to verify Summit V100 and Frontier MI-100/MI-200 GPUs.

using ROCM

Load up the AMD compiler on Spock:

module load rocm
module load craype-accel-amd-gfx908

Note both modules must be loaded for GPU offload to work.

The compiler flags for enabling GPU offload for OpenMP is the following:

CFLAGS = -fopenmp -target x86_64-pc-linux-gnu 		\
			-fopenmp-targets=amdgcn-amd-amdhsa   	\
			-Xopenmp-target=amdgcn-amd-amdhsa    	\
			-march=gfx908

To verify:

case 1

./mt-dgemm 10000

About

A collection of GEMM benchmarks to verify Summit V100 and Frontier MI-100/MI-200 GPUs.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published