clm_9703db4f30bc4049ad7e79a453b16b9a_r1

IX2-0075

① SA Source

• Source: 開啟完整 SA 文章
• Section: Prefill and Decode Phases
• Line hint: 103

Context Before

The Cost/Perf per TCO vs Interactivity/End-to-End Latency curve mostly follows the Throughput vs Interactivity/End-to-End Latency Curve: More tokens/hour leads to a lower cost per token as fixed $/hour costs are amortized over more tokens produced.

Prefill and Decode Phases

Evidence

Each forward pass loads the entire KV cache for a sequence from HBM, while only performing the computation for a single token, making decode memory (bandwidth) intensive

Context After

When prefill and decode performed on the same engine, prefill constantly disrupts decode batches leading to worse overall performance.

Disaggregated Prefill

② Atomic Claim

每次 forward pass 都必須從 HBM 載入整段 sequence 的 KV cache，但只計算一個 token，因此 decode 屬於記憶體頻寬密集型工作。

• Epistemic Mode: ASSERTED
• Mapping Status: COMPLETE

③ Semantic Frame

{
  "attribute": "BANDWIDTH",
  "context_nodes": [
    {
      "id": "04_knowledge_base/KV cache",
      "label": "KV cache"
    },
    {
      "id": "04_knowledge_base/Decode",
      "label": "decode"
    }
  ],
  "entity": {
    "id": "04_knowledge_base/HBM",
    "label": "HBM"
  },
  "frame_type": "ATTRIBUTE",
  "qualifiers": {
    "condition_text": null,
    "numeric_mentions": [],
    "temporal_mentions": []
  },
  "value": {
    "numeric_mentions": [],
    "value_text": "每次 forward pass 都必須從 HBM 載入整段 sequence 的 KV cache，但只計算一個 token，因此 decode 屬於記憶體頻寬密集型工作。"
  }
}

④ Canonical Entity Mapping

Role	Surface Label	Canonical Target
entity	HBM	HBM
context_0	KV cache	04_knowledge_base/KV cache
context_1	decode	Decode

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。