We learn how to build nVidia GPU-based (super)computers, network and program them for HPC (High Performance Computing) using CUDA C language. We apply all this to do some astrophysical simulations related to forming planetary systems.
Wednesday, May 26, 2010
[from Robert] cudak4 & cudak5 up and running
Both machines are running Fedora 12 along with the newest 256.25 Beta drivers from NVIDIA finally. The CUDA compiler is also running, and I have sent everyone an email on how to access c5 (for now). Try it out!
No comments:
Post a Comment