Re: Exclusions and CUDA

From: Axel Kohlmeyer (akohlmey_at_cmm.chem.upenn.edu)
Date: Thu Jun 04 2009 - 09:03:24 CDT

On Thu, 2009-06-04 at 10:40 +0200, Benjamin Bouvier wrote:
> Hi all,
>
> I've just compiled 2.7beta1 with CUDA support, tried to run a test
> simulation, got 'found xx exclusions but expected yy' errors, and found
> out that 2.7beta1 does not fully support exclusions.
>
> Has that changed in CVS? If not, any ETA?

yes. it is not perfect, though. from the cvs log:

2009-05-22 04:45 jim

        * src/: ComputeNonbondedBase.h, ComputeNonbondedBase2.h,
        ComputeNonbondedCUDA.C, ComputeNonbondedCUDAExcl.C,
        ComputeNonbondedPair.C, ComputeNonbondedSelf.C:

        CUDA-enabled NAMD now works with PME, but not multiple
timestepping.

2009-05-19 01:10 jim

        * src/: ComputeMap.h, ComputeMgr.C, ComputeNonbondedBase.h,
        ComputeNonbondedCUDAExcl.C, ComputeNonbondedCUDAExcl.h,
        ComputeNonbondedCUDAExcl.inl, ComputeNonbondedUtil.h,
        LdbCoordinator.C, Molecule.h, WorkDistrib.C, WorkDistrib.ci,
        WorkDistrib.h:

        Implement modified 1-4 exclusions for CUDA. Only works for
cutoff
        and pressure is still not reported correctly for CUDA
calculation.
        

>

-- 
=======================================================================
Axel Kohlmeyer   akohlmey_at_cmm.chem.upenn.edu   http://www.cmm.upenn.edu
   Center for Molecular Modeling   --   University of Pennsylvania
Department of Chemistry, 231 S.34th Street, Philadelphia, PA 19104-6323
tel: 1-215-898-1582,  fax: 1-215-573-6233,  office-tel: 1-215-898-5425
=======================================================================
If you make something idiot-proof, the universe creates a better idiot.

This archive was generated by hypermail 2.1.6 : Wed Feb 29 2012 - 15:52:53 CST