IX2-0142
① SA Source
- Source: 開啟完整 SA 文章
- Section:
DeepSeek Disagg +WideEP Results Deep Dive - Line hint:
224
Context Before
Source: SemiAnalysis InferenceX ↗
When composing disagg prefill+wideEP with FP4 on the MI355X, we observe suffers subpar performance.
Evidence
Although theoretical modeling shows that disagg inference on MI355Xs should perform way better than single node, disagg actually performs worse for higher interactivity levels due to a lack of kernel and collective optimization in the ROCm software stack when composing multiple SOTA inference optimizations together
Context After
Source: SemiAnalysis InferenceX ↗
② Atomic Claim
雖然理論模型顯示 MI355X 的 disagg inference 應明顯優於 single-node，但在較高 interactivity 下反而更慢，原因是 ROCm software stack 在同時組合多種 SOTA inference optimizations together 時，缺乏足夠的 kernel 與 collective 最佳化。
- Epistemic Mode:
ASSERTED - Mapping Status:
COMPLETE
③ Semantic Frame
{
"comparison_expression": "雖然理論模型顯示 MI355X 的 disagg inference 應明顯優於 single-node，但在較高 interactivity 下反而更慢，原因是 ROCm software stack 在同時組合多種 SOTA inference optimizations together 時，缺乏足夠的 kernel 與 collective 最佳化。",
"entities": [
{
"id": "04_knowledge_base/MI355X",
"label": "MI355X"
},
{
"id": "04_knowledge_base/ROCm",
"label": "ROCm software stack"
}
],
"frame_type": "COMPARISON",
"metric": "INTERACTIVITY",
"operator": "OUTPERFORMS",
"qualifiers": {
"condition_text": null,
"numeric_mentions": [],
"temporal_mentions": []
}
}④ Canonical Entity Mapping
| Role | Surface Label | Canonical Target |
|---|---|---|
| comparison_entity_0 | MI355X | MI355X |
| comparison_entity_1 | ROCm software stack | ROCm |
⑤ Human Review
請在 Properties 逐項確認：
- 原文 → Atomic Claim 是否忠實
- Atomic Claim → Semantic Frame 是否忠實
- Canonical Entity mapping 是否正確
- Epistemic mode 是否保留原文語氣
- 最後選擇
review_action
Review state
Markdown 內文不是正式 approval。只有 Apply bridge 寫入的 Decision Ledger event 才是正式決策。