[gmx-developers] Gromacs hardware
spoel at xray.bmc.uu.se
Mon Apr 25 22:17:41 CEST 2005
On Mon, 2005-04-25 at 13:27 -0700, Brendan Moran wrote:
> Brendan Moran wrote:
> > The benefit is parallelism, and preprogramming. The opteron is going
> > to be doing a lot of pushing variables around, and that takes time.
> > An FPGA based processor might be able to handle that portion better,
> > though it will never compare with the raw power available from a true
> > CPU.
> I should point out that I never thought that the coprocessor would be
> doing all the work. It would just be a tool to operate the most time
> consuming part of the gmx algorithm, as outlined by Erik
I don't want to discourage you, but you are probably aware of what the
competition looks like: I'm about to order some new Dual Opteron boxes,
with Dual Core chips. That'll give me 4 chips @ 2GHz in a 1U box. Such a
box would cost roughly $4000, depending on clock speed...
Assuming 1 FPGA does 1 flop per clock, you'd need roughly 15 of them to
be competitive with the 4 way Opteron box. So let's say you have the
design ready and it can go into production. How much would it cost to
produce a board with those? How many can you fit in on a regular PCI
You may also want to check the following hardware implementation:
> Brendan Moran
> gmx-developers mailing list
> gmx-developers at gromacs.org
> Please don't post (un)subscribe requests to the list. Use the
> www interface or send it to gmx-developers-request at gromacs.org.
David van der Spoel, PhD, Assoc. Prof., Molecular Biophysics group,
Dept. of Cell and Molecular Biology, Uppsala University.
Husargatan 3, Box 596, 75124 Uppsala, Sweden
phone: 46 18 471 4205 fax: 46 18 511 755
spoel at xray.bmc.uu.se spoel at gromacs.org http://xray.bmc.uu.se/~spoel
More information about the gromacs.org_gmx-developers