This is a collection of portable C benchmark implementations for the HPC dwarfs algorithmic classes as initially proposed by the report The Landscape of Parallel Computing Research: A View from Berkeley. Each benchmark is designed for performance engineering: compact, self-contained, and focused on a single computational pattern.
| Dwarf | Benchmark |
|---|---|
| Sparse Linear Algebra | SparseBench |
| N-Body Methods | MD-Bench |
| Structured Grids + Sparse Linear Algebra | CFD-Bench |
| Structured Grids | LBM-Bench |
| Spectral Methods | DFT-Bench (coming soon) |
| Signal Processing | RabbitCT |
| Streaming Kernels | TheBandwidthBenchmark |
This is an effort by NHR@FAU
