[gmx-users] hyperthreading

[Berk Hess]

Hi,

In Gromacs 4.5 there is no difference, since it does not use real thread parallelization.
Gromacs 4.5 has a built-in threaded MPI library, but openmpi also has an efficient
MPI implementation for shared memory machines. But even with proper thread
parallelization I expect the same 15 to 20% performance improvement.
(I guess the lower 8% number is because of loss when going from 8 to 16 processes
for mdrun for that particular system)

Berk

> Date: Mon, 9 Aug 2010 11:36:35 -0400
> From: chris.neale at utoronto.ca
> To: gmx-users at gromacs.org
> Subject: [gmx-users] hyperthreading
> 
> I haven't tried mdrun -nt based hyperthreading, but I have tested  
> using -np 16 on an 8 core box. I get an 8% to 18% performance increase  
> when using -np 16 and optimizing -npme as compared to -np 8 and  
> optimizing -npme. This is on a cluster of Intel Xeon E5540 aka  
> "Nehalem" with 2 quad cores in a single box and using gromacs 4.0.7. I  
> now regularly overload the number of processes.
> 
> selected examples:
> System A with 250,000 atoms:
> mdrun -np 8  -npme -1    1.15 ns/day
> mdrun -np 8  -npme  2    1.02 ns/day
> mdrun -np 16 -npme  2    0.99 ns/day
> mdrun -np 16 -npme  4    1.36 ns/day <-- 118 % performance vs 1.15 ns/day
> mdrun -np 15 -npme  3    1.32 ns/day
> 
> System B with 35,000 atoms (4 fs timestep):
> mdrun -np 8  -npme -1    22.66 ns/day
> mdrun -np 8  -npme  2    23.06 ns/day
> mdrun -np 16 -npme -1    22.69 ns/day
> mdrun -np 16 -npme  4    24.90 ns/day <-- 108 % performance vs 23.06 ns/day
> mdrun -np 56 -npme 16    14.15 ns/day
> 
> Cutoffs and timesteps differ between these runs, but both use PME and  
> explicit water.
> 
> I'd be interested in hearing about any comparisons between -np based  
> process overloading and -nt based hyperthreading.
> 
> Hope it helps,
> Chris.
> 
> -- original message --
> 
> Hi,
> 
> These are nehalem Xeons I presume?
> Then you get 15 to 20% more performance in Gromacs running 2 vs 1  
> thread or process per physical core.
> 
> Berk
> 
> > Date: Fri, 6 Aug 2010 09:24:11 -0500
> > From: dmobley at gmail.com
> > To: gmx-users at gromacs.org
> > Subject: [gmx-users] hyperthreading
> >
> > Dear All,
> >
> > I'm putting together a new Dell Xeon cluster running ROCKS 5.3 which
> > uses CENTOS (6 I believe). This is currently ~20 dual quad-cores with
> > roughly 16 GB of RAM each.
> >
> > In any case, I wanted to inquire about hyperthreading. Does anyone
> > have experience on similar machines with vs. without hyperthreading?
> > The ROCKS users list suggests that hyperthreading ought always be off
> > for HPC applications, which sounds overly simplistic to me, though I
> > more or less follow the logic of this.
> >
> > So, has anyone done any benchmarking yet in a similar setting, and
> > what thoughts do you have? I obviously can do some benchmarking myself
> > as well but I thought I'd check in with the list first.
> >
> > Thanks so much,
> > David
> >
> >
> > -- 
> > David Mobley
> > dmobley at gmail.com
> > 504-383-3662
> 
> 
> -- 
> gmx-users mailing list    gmx-users at gromacs.org
> http://lists.gromacs.org/mailman/listinfo/gmx-users
> Please search the archive at http://www.gromacs.org/search before posting!
> Please don't post (un)subscribe requests to the list. Use the 
> www interface or send it to gmx-users-request at gromacs.org.
> Can't post? Read http://www.gromacs.org/mailing_lists/users.php
 		 	   		  
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://maillist.sys.kth.se/pipermail/gromacs.org_gmx-users/attachments/20100809/c91285b0/attachment.html>