From d60d658f5f0f79c8853976aa83049be7174c4192 Mon Sep 17 00:00:00 2001 From: Xiaoyu Zhang <1182563586@qq.com> Date: Mon, 31 Aug 2026 21:31:04 +0800 Subject: [PATCH] [Kernel] Add GB300 Triton MoE configs for GLM-4.5 FP8 (#37159) Co-authored-by: Song Bian --- ...dtype=fp8_w8a8,per_channel_quant=True.json | 58 +++++++++++++++++++ 1 file changed, 58 insertions(+) create mode 100644 python/sglang/srt/layers/moe/moe_runner/triton_utils/configs/triton_3_7_1/E=161,N=192,device_name=NVIDIA_GB300,dtype=fp8_w8a8,per_channel_quant=True.json diff --git a/python/sglang/srt/layers/moe/moe_runner/triton_utils/configs/triton_3_7_1/E=161,N=192,device_name=NVIDIA_GB300,dtype=fp8_w8a8,per_channel_quant=True.json b/python/sglang/srt/layers/moe/moe_runner/triton_utils/configs/triton_3_7_1/E=161,N=192,device_name=NVIDIA_GB300,dtype=fp8_w8a8,per_channel_quant=True.json new file mode 100644 index 000000000..685592977 --- /dev/null +++ b/python/sglang/srt/layers/moe/moe_runner/triton_utils/configs/triton_3_7_1/E=161,N=192,device_name=NVIDIA_GB300,dtype=fp8_w8a8,per_channel_quant=True.json @@ -0,0 +1,58 @@ +{ + "1": { + "BLOCK_SIZE_M": 16, + "BLOCK_SIZE_N": 32, + "BLOCK_SIZE_K": 64, + "GROUP_SIZE_M": 1, + "num_warps": 4, + "num_stages": 3 + }, + "16": { + "BLOCK_SIZE_M": 16, + "BLOCK_SIZE_N": 128, + "BLOCK_SIZE_K": 128, + "GROUP_SIZE_M": 1, + "num_warps": 4, + "num_stages": 4 + }, + "256": { + "BLOCK_SIZE_M": 16, + "BLOCK_SIZE_N": 128, + "BLOCK_SIZE_K": 64, + "GROUP_SIZE_M": 32, + "num_warps": 4, + "num_stages": 3 + }, + "512": { + "BLOCK_SIZE_M": 64, + "BLOCK_SIZE_N": 128, + "BLOCK_SIZE_K": 64, + "GROUP_SIZE_M": 32, + "num_warps": 4, + "num_stages": 3 + }, + "1024": { + "BLOCK_SIZE_M": 64, + "BLOCK_SIZE_N": 128, + "BLOCK_SIZE_K": 64, + "GROUP_SIZE_M": 32, + "num_warps": 4, + "num_stages": 3 + }, + "4096": { + "BLOCK_SIZE_M": 64, + "BLOCK_SIZE_N": 128, + "BLOCK_SIZE_K": 64, + "GROUP_SIZE_M": 32, + "num_warps": 4, + "num_stages": 3 + }, + "16384": { + "BLOCK_SIZE_M": 64, + "BLOCK_SIZE_N": 128, + "BLOCK_SIZE_K": 64, + "GROUP_SIZE_M": 32, + "num_warps": 4, + "num_stages": 3 + } +}