From 1171467e91b3a1f334ad99df934ce35b9fb7255c Mon Sep 17 00:00:00 2001 From: "Li, Jiang" Date: Thu, 9 Jul 2026 15:09:31 +0800 Subject: [PATCH] [CPU] Fix Qwen-Next SSM type for AMX GDN (#48073) Signed-off-by: jiang1.li --- vllm/platforms/cpu.py | 8 ++++++++ 1 file changed, 8 insertions(+) diff --git a/vllm/platforms/cpu.py b/vllm/platforms/cpu.py index 571a8c9c2c..ed7cec519a 100644 --- a/vllm/platforms/cpu.py +++ b/vllm/platforms/cpu.py @@ -125,6 +125,14 @@ class CpuPlatform(Platform): "otherwise the performance is not optimized." ) + # AMX GDN requires float32 state + if ( + torch.cpu._is_amx_tile_supported() + and cache_config.mamba_ssm_cache_dtype != "float32" + ): + cache_config.mamba_ssm_cache_dtype = "float32" + logger.warning("Reset SSM cache type to float32 for AMX mamba attention.") + # Lagecy setting env_key = "VLLM_CPU_KVCACHE_SPACE" if env_key in os.environ and os.environ[env_key] != "":