GPU - Graphics Processing Unit computing
The world of high performance and scientific computing has changed rapidly in recent years.
To provide scientists with the latest tools and technologies, we have implemented a modern computation cluster based on GPU technology. This technology allows for the placement of a large number of processing cores on a single piece of silicon. A single GPU chip can hold more than 400 cores, as compared to a CPU chip's 6 or less.
Use of this technology is suitable for the following scientific situations:
- CPU-bounded rather than I/O-bounded JOBS
- Algorithms which can be parallelized
- When a single task's calculation time is relatively long
Around the world today there are already several hundred academic institutions performing research based on this technology including Harvard University, Cambridge, Tokyo Institute of Technology and others.
We would like to locate Institute laboratories with computational needs that are suitable for GPU processor runs, to examine the possibility of significant improvement in their job's run time duration. We are certain that this technology will help scientists deal with today's computational challenges.
MUMmerGPU: High-through DNA sequence alignment using GPUs
Accelerating HMMER using GPUs Scalable Informatics
Nvidia Corporation GPU links:
Tesla-supercomputing-solutions
Bioinformatics and Life Sciences