VR2R-0411_r1

VR2R-0411

① SA Source

• Source: 開啟完整 SA 文章
• Section: Compute Tray: Power Delivery
• Line hint: 340

Context Before

Power sloshing between Vera and Rubin is still present, a feature carried over from GB300. This enables more efficient power planning by sharing the 4800W provided amongst both GPU and CPU.

Evidence

In GPU demanding loads, 2300W goes to each GPU leaving 200W for the CPU.

Context After

Vera can then dynamically boost to higher power when GPU demand drops to help minimize GPU idle time while not overprovisioning power.

② Atomic Claim

在 GPU demand 高的負載下，每顆 GPU 分配 2,300W，剩下 200W 給 CPU。

• Epistemic Mode: ASSERTED
• Mapping Status: COMPLETE

③ Semantic Frame

{
  "frame_type": "NARY_RELATION",
  "participants": [
    {
      "node": {
        "id": "04_knowledge_base/GPU",
        "label": "GPU"
      },
      "role": "accelerator"
    },
    {
      "node": {
        "id": "04_knowledge_base/CPU",
        "label": "CPU"
      },
      "role": "cpu"
    }
  ],
  "qualifiers": {
    "condition_text": "GPU-demanding load",
    "numeric_mentions": [
      "2300W",
      "200W"
    ],
    "temporal_mentions": []
  },
  "relation_type": "GPU_HEAVY_POWER_ALLOCATION"
}

④ Canonical Entity Mapping

Role	Surface Label	Canonical Target
accelerator	GPU	GPU
cpu	CPU	CPU

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。