IX2-0084
① SA Source
- Source: 開啟完整 SA 文章
- Section:
Tensor Parallel, Expert Parallel, Data Parallel (TP, EP, DP) - Line hint:
113
Context Before
Disaggregated prefill (aka PD disaggregation or simply “disagg”) is the practice of separating the prefill and decode phases across separate pools of GPUs or clusters. These separate prefill and decode pools can be tuned independently and scaled to match the needs of workloads.
Tensor Parallel, Expert Parallel, Data Parallel (TP, EP, DP)
Evidence
DP replicates the entire model (or just parts of a model, like attention) on multiple groups of GPUs (ranks) and then load balances requests among ranks
Context After
Tracking Improvements Over Time
One of the main goals of InferenceX is to visualize performance improvements over time. While new chips are released on an O(yearly) cadence, software releases happen on an O(weekly) cadence. Our goal is to constantly update recipes with the latest and greatest software improvements and benchmark the configurations.
② Atomic Claim
DP 會在多組 GPUs（ranks）上複製整個模型，或只複製模型部分（例如 attention），再把 requests 在各 ranks 之間做 load balancing。
- Epistemic Mode:
ASSERTED - Mapping Status:
COMPLETE
③ Semantic Frame
{
"attribute": "COUNT",
"context_nodes": [],
"entity": {
"id": "04_knowledge_base/GPU",
"label": "GPUs"
},
"frame_type": "ATTRIBUTE",
"qualifiers": {
"condition_text": null,
"numeric_mentions": [],
"temporal_mentions": []
},
"value": {
"numeric_mentions": [],
"value_text": "DP 會在多組 GPUs（ranks）上複製整個模型，或只複製模型部分（例如 attention），再把 requests 在各 ranks 之間做 load balancing。"
}
}④ Canonical Entity Mapping
| Role | Surface Label | Canonical Target |
|---|---|---|
| entity | GPUs | GPU |
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。