diff options
Diffstat (limited to 'README.md')
| -rw-r--r-- | README.md | 6 |
1 files changed, 2 insertions, 4 deletions
| @@ -93,18 +93,16 @@ make tests | |||
| 93 | #### Known Working | 93 | #### Known Working |
| 94 | 94 | ||
| 95 | - NVIDIA GPUs from compute capability 3.5 through 8.9, including embedded "Jetson" GPUs | 95 | - NVIDIA GPUs from compute capability 3.5 through 8.9, including embedded "Jetson" GPUs |
| 96 | - CUDA 8.0 through 12.6 | 96 | - CUDA 6.5 through 12.6 |
| 97 | - `x86_64` and Jetson `aarch64` platforms | 97 | - `x86_64` and Jetson `aarch64` platforms |
| 98 | 98 | ||
| 99 | #### Known Issues | 99 | #### Known Issues |
| 100 | 100 | ||
| 101 | - `next_mask` will not override `stream_mask` on CUDA 11.0+ | ||
| 102 | - _As of Feb 2024, a fix for this is coming soon..._ | ||
| 103 | - `global_mask` and `next_mask` cannot disable TPCs with IDs above 128 | 101 | - `global_mask` and `next_mask` cannot disable TPCs with IDs above 128 |
| 104 | - Only relevant on GPUs with over 128 TPCs, such as the RTX 6000 Ada | 102 | - Only relevant on GPUs with over 128 TPCs, such as the RTX 6000 Ada |
| 105 | - Untested on H100 (compute capability 9.0) | ||
| 106 | - Untested on non-Jetson `aarch64` platforms | 103 | - Untested on non-Jetson `aarch64` platforms |
| 107 | - Untested on CUDA 11.8, 12.0, and 12.1 on Jetson `aarch64` | 104 | - Untested on CUDA 11.8, 12.0, and 12.1 on Jetson `aarch64` |
| 105 | - Mask bit indexes do not directly correlate to software-visible TPC/SM IDs in V4 TMD/QMDs (Hopper+; compute capability 9.0). The mask bit indexes instead appear to correspond to on-chip-units, including disabled ones; i.e. the set of pre-SM-ID-remapping and pre-floorsweeping TPCs | ||
| 108 | 106 | ||
| 109 | ## Important Limitations | 107 | ## Important Limitations |
| 110 | 108 | ||