| Name: | clpeak |
|---|---|
| Version: | 3.0.3 |
| Release: | 2.el10_2 |
| Architecture: | aarch64 |
| Group: | Unspecified |
| Size: | 2182802 |
| License: | Apache-2.0 |
| RPM: | clpeak-3.0.3-2.el10_2.aarch64.rpm |
| Source RPM: | clpeak-3.0.3-2.el10_2.src.rpm |
| Build Date: | Fri Oct 30 2026 |
| Build Host: | build-ol10-aarch64.oracle.com |
| Vendor: | Oracle America |
| URL: | https://github.com/krrishnarraj/clpeak |
| Summary: | Measure the peak achievable performance of GPU compute devices |
| Description: | A synthetic micro-benchmark that measures the peak achievable performance of GPU compute devices. It exercises tight vector / MAD / MMA loops and vendor-SDK GEMM libraries (cuBLASLt on NVIDIA, MPS on Apple) to expose what the hardware is capable of — from raw ALU peaks to near-vendor-advertised matrix throughput. clpeak began as an OpenCL-only tool. It now ships four interchangeable backends — OpenCL, Vulkan, CUDA, and Metal — running back-to-back on the same hardware, so cross-stack differences (driver lowering, instruction scheduling, extension exposure) become visible alongside the raw peak numbers. |
- Fix EL10 build compatibility
- Update to 3.0.3 upstream release - Resolves: rhbz#2558290
- Update to 3.0.2 upstream release - Resolves: rhbz#2541794
- Update to 3.0.1 upstream release - Resolves: rhbz#2541794
- drop ix86 support
- Update to 2.1.4 upstream release - Resolves: rhbz#2525161
- enable ROCm support on x86_64
- Update to 2.1.3 upstream release - Resolves: rhbz#2521842
- Update to 2.1.2 upstream release - Resolves: rhbz#2520629
- use version macro to fix the --version output issue