Projects

gpuocelot

OverviewOcelot is a dynamic compilation framework for heterogeneous systems, accomplishing this by providing various backend targets for CUDA programs. Ocelot currently allows CUDA programs to be executed on NVIDIA GPUs and x86-CPUs at full speed without recompilation. NewsDecember 15, 2009: ... [More] Preparing Ocelot 1.0.265 release. Support for native GPU and CPU execution. Currently this is a subversion-only release, you need to checkout and compile ocelot to use the CPU backend. ContributingComplete a Feature RequestIf you would like to contribute to this project and help with any of the directions on our roadmap you can do the following: Pull a task from the list of issues Implement it on your own Post a patch to the relevant issue If it is accepted we will merge it into the main codebase Ask us about becoming a registered developer Branch Our CodeIf you want to work on something not on our roadmap, but want to host your code on this site, contact us about becoming a developer and creating a branch. Start A New ProjectIf you want to work independently using Ocelot as a starting point, feel free to copy our most current release and use it internally. Special ThanksWe would like to thank the following people, who have contributed novel ideas, software, and tests to the project: Nathan Bell Sylvain Collange David Luebke Ryuta Suzuki Steve Worley [Less]

g-crossbit

What is G-CrossbitG-CrossBit is a virtual machine target to GPU-architecture based on Crossbit.We are working on strategy that makes use of GPU to improve the performance implicitly,opposite to CUDA as an explicit model. MotivationGPU has been an important computing platform for computing task ... [More] with a great arithmetic intensity,and NVIDIA as well as ATI has offered their computing model,CUDA and Brook+, for this challenge.All these models are explicit which means that programmers point out which part of code should be executed by GPU,it's not exactly the same with the original linear programming environment so as to bring some troubles for redeveloping the software with these environments. We are trying to make use of GPU implicitly through a virtual machine,G-Crossbit, which will translate the part of source code which is adapt to be execute by GPU from LLVM IR to PTX codes. Current StatusAug 1st,2009 : Designed the complete framework for translator module July 31th,2009: A translator from a subset of LLVM ISA to PTX code finished July 17th,2009: Device Driver Caller finished July 15th,2009: Linker for multi-files finished July 13th,2009: Disassembling module finished [Less]