Thank you everyone for your help. I was able to resolve the issue. This is what I found:
The main library path for cuda was already included in my LB_LIBRARAY_PATH variable (CUDA_HOME/lib64). In that directory, all of the cuda library files were there EXCEPT for the libcuda.so file. Once I figured that out, I created links in the /usr/local/cuda/lib64 for libcuda.so files.
FYI, the libcuda files were located in the directory /usr/lib64/nvidia. Be aware that the Nvidia cuda installation was a standard install (i.e. I didn’t specify unique locations). I used version 8.0 of Nvidia cuda. I don’t know if this was unique to my system, or if they changed the location of the libcuda file from ver 7.5 to 8.
Thank you again,
Alex
On Jun 16, 2017, at 11:00 AM, David Case <david.case.rutgers.edu<mailto:david.case.rutgers.edu>> wrote:
On Fri, Jun 16, 2017, Alexander Martin Spring-Connell wrote:
-L/usr/local/cuda/lib64 -L/usr/local/cuda/lib -lcublas -lcusparse
-lcudart -lcuda -lstdc++ -I/usr/local/cuda/include \
../lib/nxtsec.o ../lib/random.o \
-L/usr/localtest/amber16/lib -larpack -llapack -lblas
/usr/bin/ld: cannot find -lcuda
The code is looking in /usr/local/cuda/{lib,lib64} to find these libraries.
If your cuda installation is not there, you need to set the CUDA_HOME
environment variable, then re-run configure.
...hope this helps.....dac
_______________________________________________
AMBER mailing list
AMBER.ambermd.org<mailto:
AMBER.ambermd.org>
https://na01.safelinks.protection.outlook.com/?url=http%3A%2F%2Flists.ambermd.org%2Fmailman%2Flistinfo%2Famber&data=02%7C01%7Caspring1%40gsu.edu%7Cf8ea2d671920477b2cf208d4b4c87d09%7C515ad73d8d5e4169895c9789dc742a70%7C0%7C0%7C636332220559394566&sdata=junyNtgc8I3Mx5LO1e0KtVI0dpGhAVeh42Wz3mcLZjE%3D&reserved=0
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Fri Jun 16 2017 - 09:00:02 PDT
