IX2-0284
① SA Source
- Source: 開啟完整 SA 文章
- Section:
AMD Composability Issue on FP4, Distributed Inferencing and Wide Expert Parallelism - Line hint:
484
Context Before
SemiAnalysis is free open source software and reader-supported. To receive new posts and support our work, consider becoming a free or paid subscriber.
Subscribed
Evidence
the theoretical speed of light modelling at SemiAnalysis and at AMD show that for FP4, disaggregated inferencing with wide expert parallelism should perform better than inference on a single node of MI355X
Context After
AMD is more than six months behind on open source distributed inferencing and wide expert parallelism and FP4 composability as shown by Nvidia and SGLang team showing off their NVFP4 performance on DeepSeek six months ago ↗.
② Atomic Claim
SemiAnalysis 與 AMD 的 theoretical speed-of-light modeling 都顯示，在 FP4 下，搭配 wide expert parallelism 的 disaggregated inference 理論上應該優於單一 MI355X node。
- Epistemic Mode:
EXPECTED - Mapping Status:
COMPLETE
③ Semantic Frame
{
"comparison_expression": "SemiAnalysis 與 AMD 的 theoretical speed-of-light modeling 都顯示，在 FP4 下，搭配 wide expert parallelism 的 disaggregated inference 理論上應該優於單一 MI355X node。",
"entities": [
{
"id": "02_companies/AMD",
"label": "AMD"
},
{
"id": "04_knowledge_base/FP4",
"label": "FP4"
},
{
"id": "04_knowledge_base/Expert Parallelism",
"label": "wide expert parallelism"
},
{
"id": "04_knowledge_base/MI355X",
"label": "MI355X"
}
],
"frame_type": "COMPARISON",
"metric": "SPEED",
"operator": "OUTPERFORMS",
"qualifiers": {
"condition_text": null,
"numeric_mentions": [],
"temporal_mentions": []
}
}④ Canonical Entity Mapping
| Role | Surface Label | Canonical Target |
|---|---|---|
| comparison_entity_0 | AMD | AMD |
| comparison_entity_1 | FP4 | FP4 |
| comparison_entity_2 | wide expert parallelism | 04_knowledge_base/Expert Parallelism |
| comparison_entity_3 | MI355X | MI355X |
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。