Numeric precision vs. accuracy of modeling

Moderators: Site Moderators, FAHC Science Team

Post Reply
cbuchner1
Posts: 14
Joined: Thu Jun 19, 2008 7:08 pm

Numeric precision vs. accuracy of modeling

Post by cbuchner1 »

Would the use of double precision arithmetic actually improve the scientific return of any given F@H work unit - or is the science return currently limited more by inherent deficiencies or simplifications in the underlying physics model, rather than numeric precision?

Are there plans to use double precision arithmertics in future GPU client iterations, say GPU V3 ?

Christian
codysluder
Posts: 1024
Joined: Sun Dec 02, 2007 12:43 pm

Re: Numeric precision vs. accuracy of modeling

Post by codysluder »

For the CPU, there are two versions of Gromacs, one single precisions and one double precision. DGromacs has been used in maybe 5% of the projects, so there must be some benefit because they won't use it unless it's needed. For the same protein, double precision runs slower since SSE can do 4 simultaneous floating operations and SSE2 can do two simultaneous doubles. Dr. Pande did answer a question about this once and he said they had revised the single precision code to obtain most of the benefits of double precision without the slower speed.

I don't know whether that applies to the GPU or not, but it might.
rbpeake
Posts: 141
Joined: Sun Jun 15, 2008 4:39 pm
Hardware configuration: Intel® Core™ 2 Duo processor E8500, dual 3.16GHz cores, 6MB L2 Cache, 1333MHz FSB (45nm); 4096MB Corsair™ XMS2 DDR2-800 RAM; 256MB eVGA™ NVIDIA® GeForce™ 8600 GT Video Card
Location: NYC Metro Area

Re: Numeric precision vs. accuracy of modeling

Post by rbpeake »

codysluder wrote:For the CPU, there are two versions of Gromacs, one single precisions and one double precision. DGromacs has been used in maybe 5% of the projects, so there must be some benefit because they won't use it unless it's needed.
What is the Gromacs 33 core, then? I am using it on one protein. It is a CPU core as well. Thanks! :)
ChelseaOilman
Posts: 1037
Joined: Sun Dec 02, 2007 3:47 pm
Location: Colorado @ 10,000 feet

Re: Numeric precision vs. accuracy of modeling

Post by ChelseaOilman »

rbpeake wrote:What is the Gromacs 33 core, then? I am using it on one protein.
http://fahwiki.net/index.php/Cores#Gromacs_33_core
Post Reply