diff --git a/python/sglang/srt/layers/moe/moe_runner/triton_utils/configs/triton_3_7_1/E=256,N=512,device_name=NVIDIA_H200.json b/python/sglang/srt/layers/moe/moe_runner/triton_utils/configs/triton_3_7_1/E=256,N=512,device_name=NVIDIA_H200.json new file mode 100644 index 000000000..0d5c45163 --- /dev/null +++ b/python/sglang/srt/layers/moe/moe_runner/triton_utils/configs/triton_3_7_1/E=256,N=512,device_name=NVIDIA_H200.json @@ -0,0 +1,24 @@ +{ + "256": { + "BLOCK_SIZE_M": 16, + "BLOCK_SIZE_N": 32, + "BLOCK_SIZE_K": 64, + "GROUP_SIZE_M": 1 + }, + "512": { + "BLOCK_SIZE_M": 32, + "BLOCK_SIZE_N": 128, + "BLOCK_SIZE_K": 128, + "GROUP_SIZE_M": 32, + "num_warps": 8, + "num_stages": 3 + }, + "1024": { + "BLOCK_SIZE_M": 64, + "BLOCK_SIZE_N": 64, + "BLOCK_SIZE_K": 64, + "GROUP_SIZE_M": 1, + "num_warps": 4, + "num_stages": 3 + } +} diff --git a/python/sglang/srt/layers/moe/moe_runner/triton_utils/configs/triton_3_7_1/E=256,N=512,device_name=NVIDIA_H200_down.json b/python/sglang/srt/layers/moe/moe_runner/triton_utils/configs/triton_3_7_1/E=256,N=512,device_name=NVIDIA_H200_down.json new file mode 100644 index 000000000..972b0b755 --- /dev/null +++ b/python/sglang/srt/layers/moe/moe_runner/triton_utils/configs/triton_3_7_1/E=256,N=512,device_name=NVIDIA_H200_down.json @@ -0,0 +1,26 @@ +{ + "256": { + "BLOCK_SIZE_M": 16, + "BLOCK_SIZE_N": 32, + "BLOCK_SIZE_K": 64, + "GROUP_SIZE_M": 1 + }, + "512": { + "BLOCK_SIZE_M": 32, + "BLOCK_SIZE_N": 64, + "BLOCK_SIZE_K": 128, + "GROUP_SIZE_M": 32, + "num_warps": 4, + "num_stages": 3, + "USE_TMA": true + }, + "1024": { + "BLOCK_SIZE_M": 64, + "BLOCK_SIZE_N": 128, + "BLOCK_SIZE_K": 64, + "GROUP_SIZE_M": 1, + "num_warps": 4, + "num_stages": 3, + "USE_TMA": true + } +}