VR2R-0692_r1

VR2R-0692

① SA Source

• Source: 開啟完整 SA 文章
• Section: VR NVL72 TCO: BoM and Power Budget Analysis
• Line hint: 601

Context Before

The VR NVL72 Component BoM and Power Budget Model provides a detailed analysis on the BoM and the Power Budget of the rack system.

Evidence

The VR NVL72 is more expensive on a per-GPU capital cost basis, ~45% higher vs GB300s and ~14-15% higher vs the MI4XX given a higher server cost on a per GPU basis.

Context After

This results in a higher Capital Cost of Ownership (TCO). For example VR NVL 72 Hyperscaler Arista has a capital cost of $3.28to[[04_{k}nowledg{e}_{b}ase/MI400|MI4XX]][[04_{k}nowledg{e}_{b}ase/Hyperscaler|Hyperscaler]]of$2.86 per hour per GPU over a 4 year useful life.

② Atomic Claim

以每顆 GPU 的 capital cost 計算，VR NVL72 約比 GB300 高 45%，也比 MI4XX 高約 14–15%；主因是每顆 GPU 對應的 server cost 較高。

• Epistemic Mode: ASSERTED
• Mapping Status: COMPLETE

③ Semantic Frame

{
  "frame_type": "NARY_RELATION",
  "participants": [
    {
      "node": {
        "id": "04_knowledge_base/VR NVL72",
        "label": "VR NVL72"
      },
      "role": "subject_platform"
    },
    {
      "node": {
        "id": "04_knowledge_base/GB300",
        "label": "GB300"
      },
      "role": "comparison_platform"
    },
    {
      "node": {
        "id": "04_knowledge_base/MI400",
        "label": "MI4XX"
      },
      "role": "comparison_platform"
    },
    {
      "node": {
        "id": "04_knowledge_base/GPU",
        "label": "GPU"
      },
      "role": "cost_unit"
    }
  ],
  "qualifiers": {
    "condition_text": "per-GPU capital cost; higher per-GPU server cost is cited as cause",
    "numeric_mentions": [
      "45%",
      "14-15%"
    ],
    "temporal_mentions": []
  },
  "relation_type": "VR_NVL72_PER_GPU_CAPEX_PREMIUM"
}

④ Canonical Entity Mapping

Role	Surface Label	Canonical Target
subject_platform	VR NVL72	04_knowledge_base/VR NVL72
comparison_platform	GB300	GB300
comparison_platform	MI4XX	MI400
cost_unit	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。