From b390e9518efaad74ff8c42e93c20d295fecc5855 Mon Sep 17 00:00:00 2001 From: "jacky.cheng" Date: Thu, 20 Aug 2026 06:05:40 +0000 Subject: [PATCH 1/2] [AMD][MI35X] Sweep TP2/EP1 instead of TP2/EP2 on the Qwen3.5 MXFP4 MI355X AgentX arm B200's qwen3.5-fp4-b200-sglang-agentic-mtp sweeps TP2/EP1, so the MI355X TP2/EP2 point was not comparable to any B200 column. Swap it for TP2/EP1 and extend the concurrency list to 32 to see where the 2-GPU arm tops out on GPU-resident KV. The TP4/EP1 arm and the benchmark script are untouched. --- configs/amd-master.yaml | 2 +- perf-changelog.yaml | 9 +++++++++ 2 files changed, 10 insertions(+), 1 deletion(-) diff --git a/configs/amd-master.yaml b/configs/amd-master.yaml index 59f655945..40836fe2c 100644 --- a/configs/amd-master.yaml +++ b/configs/amd-master.yaml @@ -372,7 +372,7 @@ qwen3.5-fp4-mi355x-sglang-agentic-mtp: agentic-coding: - dram-utilization: 0.80 search-space: - - { tp: 2, ep: 2, spec-decoding: mtp, kv-offloading: none, conc-list: [1, 4, 8, 12, 16, 20] } + - { tp: 2, ep: 1, spec-decoding: mtp, kv-offloading: none, conc-list: [1, 4, 8, 12, 16, 20, 24, 28, 32] } - { tp: 4, ep: 1, spec-decoding: mtp, kv-offloading: none, conc-list: [1, 4, 8, 12, 16, 20, 24, 28, 32, 40] } qwen3.5-fp4-mi355x-sglang-disagg: diff --git a/perf-changelog.yaml b/perf-changelog.yaml index 93b6a7c1b..a585be2ea 100644 --- a/perf-changelog.yaml +++ b/perf-changelog.yaml @@ -6251,3 +6251,12 @@ description: - "Refresh with lower stream interval to collect correct client metrics" pr-link: https://github.com/SemiAnalysisAI/InferenceX/pull/2686 + +- config-keys: + - qwen3.5-fp4-mi355x-sglang-agentic-mtp + scenario-type: + - agentic-coding + description: + - "Replace the TP2/EP2 arm with TP2/EP1 so the 2-GPU point matches qwen3.5-fp4-b200-sglang-agentic-mtp." + - "Extend the TP2 concurrency list to 1, 4, 8, 12, 16, 20, 24, 28, and 32." + pr-link: https://github.com/SemiAnalysisAI/InferenceX/pull/2693 From 5c9797c65d461504b31736dd39960eb35ddf8d13 Mon Sep 17 00:00:00 2001 From: "jacky.cheng" Date: Thu, 20 Aug 2026 06:06:09 +0000 Subject: [PATCH 2/2] [AMD][MI35X] Point the Qwen3.5 AgentX TP2/EP1 changelog entry to PR #2694 --- perf-changelog.yaml | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/perf-changelog.yaml b/perf-changelog.yaml index a585be2ea..d98491b1c 100644 --- a/perf-changelog.yaml +++ b/perf-changelog.yaml @@ -6259,4 +6259,4 @@ description: - "Replace the TP2/EP2 arm with TP2/EP1 so the 2-GPU point matches qwen3.5-fp4-b200-sglang-agentic-mtp." - "Extend the TP2 concurrency list to 1, 4, 8, 12, 16, 20, 24, 28, and 32." - pr-link: https://github.com/SemiAnalysisAI/InferenceX/pull/2693 + pr-link: https://github.com/SemiAnalysisAI/InferenceX/pull/2694