Abstract
Graphics processing units (GPUs) can provide excellent speedups on some, but not all, general-purpose workloads. Using a set of computational GPU kernels as examples, the authors show how to adapt kernels to utilize the architectural features of a GeForce 8800 GPU and what finally limits the achievable performance.
| Original language | English (US) |
|---|---|
| Article number | 4814979 |
| Pages (from-to) | 16-26 |
| Number of pages | 11 |
| Journal | Computing in Science and Engineering |
| Volume | 11 |
| Issue number | 3 |
| DOIs | |
| State | Published - May 2009 |
Keywords
- Benchmarks
- CUDA
- Compute unified device architecture
- Computer architecture
- GPGPU
- General-purpose computing on GPU
- Software optimization
ASJC Scopus subject areas
- Computer Science(all)
- Engineering(all)