From: Madhulika Gupta (gmadhulika1909_at_gmail.com)
Date: Wed Nov 18 2015 - 02:47:38 CST
Dear All,
I am trying to compile NAMD2.11 Source code on the central computing
facility at our institute and I get the error as listed below. I have tried
compiling both MPI version and Infiniband version but both result in the
same error. The earlier versions worked just fine on the same machine.
./config Linux-x86_64-g++ --charm-arch mpi-linux-x86_64 --with-cuda
--cuda-prefix /home/soft/cuda-7.0/
Selected arch file arch/Linux-x86_64-g++.arch contains:
NAMD_ARCH = Linux-x86_64
CHARMARCH = multicore-linux64
CXX = g++ -m64 -std=c++0x -O3
CXXOPTS = -fexpensive-optimizations -ffast-math
CC = gcc -m64 -O3
COPTS = -fexpensive-optimizations -ffast-math
ERROR: CUDA builds require non-MPI SMP or multicore Charm++ arch for
reasonable performance.
Consider ibverbs-smp or verbs-smp (InfiniBand), gni-smp (Cray), or
multicore (single node).
-------------------------------------------------------------------------
I have also tried using namd binaries under the following
- Linux-x86_64-multicore-CUDA
<http://www.ks.uiuc.edu/Development/Download/download.cgi?UserID=&AccessCode=&ArchiveID=1357>
(NVIDIA CUDA acceleration)
- Linux-x86_64-ibverbs-smp-CUDA
<http://www.ks.uiuc.edu/Development/Download/download.cgi?UserID=&AccessCode=&ArchiveID=1356>
(NVIDIA CUDA with InfiniBand)
but they result in the following error:
Exclusions: detected 16049 expected 16709
FATAL ERROR: Incorrect number of exclusions detected
Please pay heed to the issue and please help me to resolve the same.
Regards
Madhulika
Research Scholar
Department of Chemistry
Indian Institute of Technology, Delhi
New Delhi-110016
This archive was generated by hypermail 2.1.6 : Thu Dec 31 2015 - 23:22:14 CST