OpenCL Client?

[Laurijan]

Hi!


Is there a F@H OpenCL client on the way? And would it be better or equally powerful to PhysX? If yes the everyone would want a 5k series card for folding right?


Lauri

[W1zzard]

i remember reading something from stanford recently about a v3 client that uses some openmm thing that supports opencl. dont remember the link though. also no info there on performance, date, availability etc.

physx isnt related to folding at all. it's is an api for games sitting on top of gpu calculations, you maybe meant cuda ?

nobody knows how anything performs on the v3 client so cant say whether you want ati or nv

[Laurijan]

Originally Posted by W1zzard

i remember reading something from stanford recently about a v3 client that uses some openmm thing that supports opencl. dont remember the link though. also no info there on performance, date, availability etc.

physx isnt related to folding at all. it's is an api for games sitting on top of gpu calculations, you maybe meant cuda ?

nobody knows how anything performs on the v3 client so cant say whether you want ati or nv

I only knew that "something" made Nvidia card superior in F@H but didnt know what exactly... always thought it was PhysX

[NastyHabits]

From the Stanford Folding@Home web-site, September 29, 2009

3) GPU3: Next generation GPU core, based on OpenMM. We have been making major advances in GPU simulation, with the key advances going into OpenMM, our open library for molecular simulation. OpenMM started with our GPU2 code as a base, but has really flourished since then. Thus, we have rewritten our GPU core to use OpenMM and we have been testing that recently as well. It is designed to be completely backward compatible, but should make simulations much more stable on the GPU as well as add new science features. A key next step for OpenMM is OpenCL support, which should allow much more efficient use of new ATI GPUs and beyond.

Bold type was added by me.

[NastyHabits]

Buried in AMD's 2010 product roadmap/announcement was this gem:

expand homegrown DirectCompute 11 and OpenCL developer tools,

Hopefully they'll send a copy to Stanford ASAP.