[AMD/ROCm] qwen3.5-fp8-mi355x-atom, Bump image to rocm/atom:rocm7.2.3_ubuntu24.04_py3.12_pytorch_release_2.10.0_atom20260511 (#1411)

* Update qwen3.5-fp8-mi355x-atom ATOM image to nightly 20260511

Bump ATOM image to rocm/atom:rocm7.2.3_ubuntu24.04_py3.12_pytorch_release_2.10.0_atom20260511.
TP=4 shows +3.2% to +16.3% throughput improvement across 1k1k and 8k1k workloads (concurrency 4-256).

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

* Update perf-changelog.yaml

* Update perf-changelog.yaml

---------

Co-authored-by: Claude Sonnet 4.6 <noreply@anthropic.com>
Co-authored-by: functionstackx <47992694+functionstackx@users.noreply.github.com>

Author: 3 people

.github/configs/amd-master.yaml

@@ -305,7 +305,7 @@ qwen3.5-fp8-mi355x-sglang-agentic:
       - { tp: 8, ep: 1, offloading: none, conc-list: [1, 2, 4, 8, 16, 32] }
 
 qwen3.5-fp8-mi355x-atom:
-  image: rocm/atom:rocm7.2.2_ubuntu24.04_py3.12_pytorch_release_2.10.0_atom0.1.2.post
+  image: rocm/atom:rocm7.2.3_ubuntu24.04_py3.12_pytorch_release_2.10.0_atom20260511
   model: Qwen/Qwen3.5-397B-A17B-FP8
   model-prefix: qwen3.5
   runner: mi355x

perf-changelog.yaml

@@ -3028,3 +3028,11 @@
   description:
     - "Add MTP/EAGLE speculative-decoding sibling for dsr1-fp4-b200-sglang (model: nvidia/DeepSeek-R1-0528-FP4-V2) on lmsysorg/sglang:v0.5.12-cu130"
   pr-link: https://github.com/SemiAnalysisAI/InferenceX/pull/1522
+
+- config-keys:
+    - qwen3.5-fp8-mi355x-atom
+  description:
+    - "Bump ATOM image to rocm/atom:rocm7.2.3_ubuntu24.04_py3.12_pytorch_release_2.10.0_atom20260511"
+    - "TP=4 shows +3.2% to +16.3% throughput improvement across 1k1k and 8k1k workloads (concurrency 4-256)"
+  pr-link: https://github.com/SemiAnalysisAI/InferenceX/pull/1411
+