Add some plans

This commit is contained in:
Chenggang Zhao 2025-03-04 15:54:46 +08:00
parent 1553fc42bf
commit 592296cd45

View File

@ -280,6 +280,13 @@ For two micro-batch overlapping, you can refer to the following figure. With our
![low-latency](figures/low-latency.png)
## Roadmap
- [ ] A100 support (intranode only)
- [ ] Support BF16 for the low-latency dispatch kernel
- [ ] Support NVLink protocol for intranode low-latency kernels
- [ ] SM-free normal kernels
## Notices
#### Easier potential overall design