GPU Automatically Tuned Linear Algebra Software
C++ Other
Switch branches/tags
Nothing to show
Clone or download
Fetching latest commit…
Cannot retrieve the latest commit at this time.
Permalink
Failed to load latest commit information.
docs
COPYING
COPYING.LESSER
GatlasAppUtil.cpp
GatlasAppUtil.hpp
GatlasBenchmark.cpp
GatlasBenchmark.hpp
GatlasCodeText.cpp
GatlasCodeText.hpp
GatlasFormatting.cpp
GatlasFormatting.hpp
GatlasOperator.cpp
GatlasOperator.hpp
GatlasQualifier.cpp
GatlasQualifier.hpp
GatlasType.cpp
GatlasType.hpp
INSTALL
KernelBaseMatmul.cpp
KernelBaseMatmul.hpp
KernelBaseMatvec.cpp
KernelBaseMatvec.hpp
KernelBaseSaxpy.cpp
KernelBaseSaxpy.hpp
KernelMatmulBuffer.hpp
KernelMatmulImage.hpp
KernelMatvecBuffer.hpp
KernelMatvecImage.hpp
KernelProbeAutoVectorize.hpp
KernelSaxpyBuffer.hpp
KernelSaxpyImage.hpp
Makefile.ATI
Makefile.NVIDIA
Makefile.common
OCLApp.cpp
OCLApp.hpp
OCLApp.tcc
OCLAppUtil.cpp
OCLAppUtil.hpp
OCLBase.cpp
OCLBase.hpp
OCLSTL.hpp
OCLUtil.cpp
OCLUtil.hpp
OCLUtil.tcc
README
benchMatmul.cpp
benchMatvec.cpp
benchSaxpy.cpp
declare_namespace
oclInfo.cpp
printMatmul.cpp
printMatvec.cpp
printSaxpy.cpp
probeAutoVectorize.cpp
purgeJournal.cpp
retry
using_namespace
wavedims

README

GATLAS - GPU Automatically Tuned Linear Algebra Software

Chris Jang
fastkor@gmail.com

October 18 2010

Download source code: http://github.com/cjang/GATLAS
Website: http://golem5.org/gatlas

Successor project: https://github.com/cjang/chai

Note: Please read INSTALL .

===============================================================================
HISTORY

May 2010 - First generation design developed for ATI Radeon 5870 GPU
Aug 2010 - Second generation with ATI 5870, 5770, 5670, 5440 and NVIDIA 480
Oct 2010 - Third generation with experimental branch merged back to master

===============================================================================
THANKS

Romain Dolbeau for testing the first generation code on a variety of platforms
including: Supermicro mainboard with dual HD 5870s; nVidia Tesla Fermi C2050;
PS3 with STI Cell. Romain gave very valuable feedback at an early stage of
development that contributed to the second generation technology roadmap.

Michael Chu and AMD/ATI for a box of engineering samples. This has been very
helpful to ensure good support across a wide range of GPU cards. It also gave
insight into kernel designs that generalize well across different GPU models.