IX2-0209
① SA Source
- Source: 開啟完整 SA 文章
- Section:
Nvidia Disagg Prefill and WideEP - Line hint:
368
Context Before
NVLink domain (within the NVL72 rack): 72 GPUs connected via NVLink with 900 GB/s uni-directional bandwidth per GPU. This is roughly 7-10x the bandwidth of the InfiniBand/Ethernet based scale-out network.
Evidence
Note that all our testing for Nvidia was conducted on InfiniBand based clusters
Context After
TP shards every layer’s weight matrices across GPUs. This means that every single token at every single layer requires up to two all-reduce communications (one after the column-parallel GEMM, one after the row-parallel GEMM). For EP, all-to-all is done only at MoE layers. Each GPU sends only the tokens routed to each expert. This means cheaper comms across all layers for EP vs TP.
Because EP’s all-to-all communication bandwidth requirements scale with the number of participants, staying within the high-bandwidth NVLink domain before having to cross the slower IB/Eth fabric is better. With NVL72, EP across 72 GPUs is possible without ever leaving NVLink, whereas previous generations (with only 8-GPU NVLink domains) could only do EP across 8 GPUs at NVLink speed before hitting the slower IB/Eth networks.
② Atomic Claim
InferenceX 對 Nvidia 的所有測試都在以 InfiniBand 為基礎的 clusters 上進行。
- Epistemic Mode:
ASSERTED - Mapping Status:
PARTIAL
③ Semantic Frame
{
"attribute": "LIFECYCLE_STATUS",
"context_nodes": [
{
"id": "04_knowledge_base/InfiniBand",
"label": "InfiniBand"
}
],
"entity": {
"id": "02_companies/NVDA",
"label": "Nvidia"
},
"frame_type": "ATTRIBUTE",
"qualifiers": {
"condition_text": null,
"numeric_mentions": [],
"temporal_mentions": []
},
"value": {
"numeric_mentions": [],
"value_text": "InferenceX 對 Nvidia 的所有測試都在以 InfiniBand 為基礎的 clusters 上進行。"
}
}④ Canonical Entity Mapping
| Role | Surface Label | Canonical Target |
|---|---|---|
| entity | Nvidia | NVDA |
| context_0 | InfiniBand | InfiniBand |
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。