| Index | NAME | Function |
|---|---|---|
| D.2.1 | GPUFL | Master flag to control what calculations are performed on the GPU |
| D.2.2 | GPUPR | Print output level |
| D.2.3 | CUPATH | Path to nvidia-smi program |
| D.2.4 | MAXGPU | Number of GPU processors to use |
| D.2.5 | GPUCPU | Bind GPU thread to CPU cores |
| D.2.6 | NOGPU | Do not use specified GPU |
| D.2.7 | IMMPCT ITRPCT ITMPCT CSAFE |
Percent of work done by CPUs |
| D.2.8 | GPUPCT | Percent of GPU free memory to use |
| D.2.9 | ICHUNK ITRCHK ITMCHK |
Transfer block size |
| D.2.10 | CGTEST CGMPCT CGLOOP |
Test CPU to GPU transfer rates |
| D.2.11 | MNCORE | Only use GPUs with MNCORE cores |
| D.2.12 | MAXEVT | Maximum recorded events on a call (read-only) |
| D.2.13 | GPUTHI GPUTLO GPUTWS |
GPU temperature to stop processing for cooling |