[Beowulf] coprocessor to do "physics calculations"
Many of your questions may have already been answered in earlier discussions or in the FAQ. The search results page will indicate current discussions as well as past list serves, articles, and papers.
Florian Haberl Florian.Haberl at chemie.uni-erlangen.deFri May 5 00:32:13 PDT 2006
- Previous message: [Beowulf] coprocessor to do "physics calculations"
- Next message: [Beowulf] coprocessor to do "physics calculations"
- Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]
Hi, On Friday 05 May 2006 03:24, Bill Broadley wrote: > On Thu, May 04, 2006 at 01:09:07PM -0700, Jim Lux wrote: > > Ran across an interesting device: > > http://physx.ageia.com/ > > > > It's some sort of coprocessor (marketed to the gaming community) that > > looks like its designed to efficiently numerically integrate equations of > > motion. Clever idea. Not much substantive info in the whitepaper. > > http://physx.ageia.com/whitepaper_avanced_gaming_physics.pdf > > Anyone know if it can handle double precision floating point? could be a nice competition to clearspeed or nec machines (md engine or something like this). It can calculate: - Density, friction and bounciness - Newtonian dynamics - Joints and springs - Small particle dynamics - Cloth - Gravity So all things around for speeding up molecular dynamics simulations: - vibrational forces (springs) - rotational forces (joints) - nonbonded forces (similar to gravity) - update position and velocity (dynamics) - update the neighbour list For me it is not clear, if i can get the results back out of the graphics card. Also there are some ongoning developments from Havok FX (250GFlops/s http://www.havok.com/ ) and something called ATI: DPP-Layer (374GFLOPS/s) The Ageia PhysX claimes to reach 20 GFLOPS/s. Greetings, Florian -- ------------------------------------------------------------------------------- Florian Haberl Computer-Chemie-Centrum Universitaet Erlangen/ Nuernberg Naegelsbachstr 25 D-91052 Erlangen Mailto: florian.haberl AT chemie.uni-erlangen.de ------------------------------------------------------------------------------- -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/pgp-signature Size: 189 bytes Desc: not available Url : http://www.scyld.com/pipermail/beowulf/attachments/20060505/ae74927c/attachment.bin
- Previous message: [Beowulf] coprocessor to do "physics calculations"
- Next message: [Beowulf] coprocessor to do "physics calculations"
- Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]
More information about the Beowulf mailing list
