SA260324R-0097_r1

SA260324R-0097

① SA Source

• Source: 開啟完整 SA 文章
• Section: GPU and LPU Integration: Attention FFN Disaggregation (AFD)
• Line hint: 105

Context Before

As we explained in our InferenceX article, LLM inference involves two phases: prefill and decode.

Evidence

Prefill processes the full input context: It is compute-intensive, which is suitable for GPUs.

Context After

On the other hand, decode predicts new tokens and is memory-bounded. Decode is latency-sensitive because the model predicts new tokens one by one, and LPU’s high SRAM bandwidth and low-latency capabilities can help accelerate this iterative process.

② Atomic Claim

Prefill 會處理完整 input context；它屬於 compute-intensive 工作，因此適合由 GPU 執行。

• Epistemic Mode: ASSERTED
• Mapping Status: COMPLETE

③ Semantic Frame

{
  "attribute": "WORKLOAD_CHARACTERISTICS",
  "context_nodes": [
    {
      "id": "04_knowledge_base/GPU",
      "label": "GPU"
    }
  ],
  "entity": {
    "id": "04_knowledge_base/Prefill",
    "label": "Prefill"
  },
  "frame_type": "ATTRIBUTE",
  "qualifiers": {
    "condition_text": null,
    "numeric_mentions": [],
    "temporal_mentions": []
  },
  "value": {
    "numeric_mentions": [],
    "value_text": "處理完整 input context，屬於 compute-intensive 工作，因此適合 GPU。"
  }
}

④ Canonical Entity Mapping

Role	Surface Label	Canonical Target
entity	Prefill	Prefill
context_node	GPU	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。