I just set up CUDA-5.5 on Ubuntu 12-04 and my GPU is the NVIDIA NVS 3100M. I have set up a page where I uploaded my experience doing it and potential applications of parallel programming using CUDA that I am looking at right now: https://sites.google.com/site/niravadesaishomepage/home/research/nvidia-gpu-programming. Any suggestions you might have, are welcome. I am looking for the CUBLAS libraries right now, esp. the libcublas.so. If any of you know where to find them, it will be very helpful.
CUBLAS is located in the lib or lib64 directory of your CUDA installation. If you have chosen the default install location, that would probably be:
/usr/local/cuda/lib64/libcublas.so
Be sure to export the lib or lib64 directory to your LD_LIBRARY_PATH to ensure your binary can find it when running your application.
Hi Lee. The folder you mentioned only has a link to a shared library file which is non existant and so I am having build error when trying to build included projects that use a CUBLAS library.
To the best of my knowledge when installed the CUDA toolkit creates a directory indicating the version (such as /usr/local/cuda-5.5) from which the /usr/local/cuda directory points to. Within the /usr/local/cuda/lib64 directory, each libXYZ.so represents a symbolic link to the version of the library installed with that release of the toolkit. For CUDA 5.5, one such example would be:
libcublas.so -> libcublas.so.5.5
libcublas.so.5.5 -> libcublas.so.5.5.22
libcublas.so.5.5.22
In this case, the shared library symlink libcublas.so requires the actual library libcublas.so.5.5.22, which is the version included with this release. If any of the intermediary files are missing, this is not going to work. Have you tried to reinstall the toolkit, and check if the required libraries are installed? If not, and you have the so.5.5.22 version, could you modify the pointer to look for this directly and attempt a build?
I see. Thanks. I have those files so I will have to figure out what's wrong. In any case I am planning to do FFT right now.
Dear Desai,
As I'm a young physics research student, your conversation on CUDA attracts me. Of course, I cannot understand it fully. Actually I'm trying to solve Nonlinear Partial Differential Equations(NPDE) numerically. Can I able to use CUDA for simulation (especially for 3D type equations)?. I find your referred site in the link ' https://sites.google.com/site/niravadesaishomepage/home/research/nvidia-gpu-programming' is useful but I'm not able to watch the videos because the firefox browser shows that the format of videos in iTunes. What can I do to watch the videos? I'm using Ubuntu 12.04, in 8 core processors workstation for solving 3D (NPDE).
- Badges
- Science topic
- Similar topics
- Human-Computer Interaction
- Collaboration