songhexiang 4dd1e68ac8 For the SMs which calculate metadata in notify_dispatch, each warp in the SM is used to calculate the metadata of one channel. The default configuration is 8 warps for 10 channels, which needs two rounds of loop. Maybe the number of warps can be configured to the number of the channels so that one loop is enough.

2025-03-28 06:43:29 +00:00

..

kernels

For the SMs which calculate metadata in notify_dispatch, each warp in the SM is used to calculate the metadata of one channel. The default configuration is 8 warps for 10 channels, which needs two rounds of loop. Maybe the number of warps can be configured to the number of the channels so that one loop is enough.

2025-03-28 06:43:29 +00:00

CMakeLists.txt

Initial commit

2025-02-25 09:07:53 +08:00

config.hpp

Support zero-copy for low-latency combine

2025-03-18 15:44:26 +08:00

deep_ep.cpp

Support zero-copy for low-latency combine

2025-03-18 15:41:50 +08:00

deep_ep.hpp

Support zero-copy for low-latency combine

2025-03-18 15:41:50 +08:00

event.hpp

Initial commit

2025-02-25 09:07:53 +08:00