SA260324R-0131_r1

SA260324R-0131

① SA Source

• Source: 開啟完整 SA 文章
• Section: Speculative Decoding
• Line hint: 167

Context Before

For LPUs, deploying a draft model or MTP layers is quite different from applying AFD. FFNs are stateless, while draft models and MTP layers require dynamic KV cache loading.

Evidence

Each FFN is around hundreds of megabytes

Context After

, whereas draft models and MTP layers take up tens of gigabytes. To support this memory usage, LPUs can access up to 256 GB of DDR5 per Fabric Expansion Logic FPGAs on the LPX compute tray.

② Atomic Claim

每個 FFN 的容量約為數百 MB。

• Epistemic Mode: ASSERTED
• Mapping Status: COMPLETE

③ Semantic Frame

{
  "attribute": "CAPACITY",
  "context_nodes": [],
  "entity": {
    "id": "04_knowledge_base/Feed-forward Network",
    "label": "FFN"
  },
  "frame_type": "ATTRIBUTE",
  "qualifiers": {
    "condition_text": null,
    "numeric_mentions": [],
    "temporal_mentions": []
  },
  "value": {
    "numeric_mentions": [],
    "value_text": "每個 FFN 的容量約為數百 MB。"
  }
}

④ Canonical Entity Mapping

Role	Surface Label	Canonical Target
entity	FFN	04_knowledge_base/Feed-forward Network

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。