DATE: Thursday, September 25, 2008
TIME: 12:00 pm - 1:00 pm
PLACE: Wean Hall 8220

Lee Ward
Sandia National Laboratories

Efficacy of GPUs in RAID Parity Calculations

RAID systems are under pressure to reconstruct data faster because of ever larger disk capacities. The parity computation is the core function of RAID controllers. Although traditionally implemented in a dedicated ASIC device, more and more storage vendors are using commodity processor resources to compute RAID. This talk looks at using graphics co-processor resources becoming available for high performance computing, specifically the NVIDIA GeForce 880 GTX CUDA, for implementing RAID without special purpose ASIC support.

Lee Ward is a principal member of technical staff at Sandia National Laboratories. As an inveterate student of operating systems and file systems, his interests have provided the opportunity to make contributions in high performance, parallel file systems, IO libraries, hierarchical storage management, and compute cluster integration/management systems. The IO performance of Sandia current largest computer was recently documented by his team:

Visitor Host:
Garth Gibson
Visitor Coordinator: Angie Miller

For Further Seminar Info:
Karen Lindenfelser (8-6716)
or visit