Tehdyt toimenpiteet
CUDA for scientific computing
Programming GPUs to accelerate algorithms and heavy computations has over the last years gained a lot of interest. This course will offer two days of CUDA training and hands-on exercises. It will also briefly cover the new OpenCL language. This course is intended for people who want to take advantage of this new technology and that has ordinary programming skills, but are new to CUDA.
Learn CUDA and write your fastest program ever!
Program
Thu Sep 10
09:00-12:15 Module I: GPU introduction and CUDA basics
12:15-13:15 Lunch
13:15-16:30 Module II: More advanced topics in CUDA
16:30- Social event
Fri Sep 11
09:00-12:15 Module III: CUDA optimization
12:15-13:15 Lunch
13:15-16:30 Module IV: OpenCL
Each module consists of lectures and related hands-on exercises, separted by a short break. Exercises are carried out with Nvidia Tesla C1060 GPGPU cards installed in the computers of CSC's training classroom. You may also carry them out using the graphics card of your own laptop, if it supports Cuda and you have the Cuda SDK installed beforehand.
Note: The scope of Module IV depends on the status of OpenCL at the time for the training. NVIDIA is currently running a beta-testing that is expected to finish well in time before the training in September. If OpenCL at that time is not a stable platform, the exercise will be replaced with a more advanced CUDA example.
Registration
Additional information
Further information is provided by Pekka Manninen (pekka.manninen(at)csc.fi, +358 50 3819 039).
