Update GLM-5 H200 FP8 (#25814)

Signed-off-by: Faradawn Yang <73060648+faradawn@users.noreply.github.com>

docs_new/cookbook/autoregressive/GLM/GLM-5.mdx

@@ -109,6 +109,7 @@ sglang serve \
   --speculative-num-steps 3 \
   --speculative-eagle-topk 1 \
   --speculative-num-draft-tokens 4 \
+  --enable-flashinfer-allreduce-fusion \
   --mem-fraction-static 0.85 \
   --host 0.0.0.0 \
   --port 30000

docs_new/src/snippets/autoregressive/glm-5-deployment.jsx

@@ -206,6 +206,11 @@ export const GLM5Deployment = () => {
       cmd += ' \\\n  --enable-flashinfer-allreduce-fusion';
     }
 
+    // H200 FP8: flashinfer allreduce fusion.
+    if (hardware === 'h200' && effectiveQuant === 'fp8') {
+      cmd += ' \\\n  --enable-flashinfer-allreduce-fusion';
+    }
+
     cmd += ` \\\n  --mem-fraction-static ${memFraction}`;
     return cmd;
   };