| 1 |
transparent schedule |
Two simple vector addition programs that run many times, demonstrating the transparent scheduling of tasks. |
CUDA, HIP, LevelZero, OpenCL |
| 2 |
give hints |
An example of showing how to create XQueue and give hints to XSched using XQueue APIs & Hint APIs |
CUDA, HIP |
| 3 |
intra-process schedule |
Using XQueue APIs & Hint & LocalScheduler for scheduling within a process |
CUDA, HIP |
| 4 |
manual schedule |
Manually scheduling within a process using XQueue APIs including Resume & Suspend |
CUDA, HIP, LevelZero |
| 5 |
inference serving |
Integrate XSched into the TensorRT backend of NVIDIA Triton Inference Server to enable priority-based scheduling of multi-model inference tasks. |
CUDA |
| 6 |
gui window |
Scheduling tasks based on the X11 window activity for GUI applications. |
LevelZero |
| 7 |
terminal window |
Scheduling tasks based on the X11 window activity for terminal applications. |
CUDA, HIP |
| 8 |
non-transparent schedule |
An example of non-transparent scheduling using XQueue APIs & submiting tasks to XQueue manually. |
CUDA, LevelZero |
| 9 |
llama.cpp with XSched |
Integrate XSched into Llama.cpp Inference Server to enable priority-based scheduling of inference tasks. |
CUDA |
