clm_a1ec2f55c8724a288df21f1850a2c2bd_r1

VR2-0637

① SA Source

• Source: 開啟完整 SA 文章
• Section: Connecting GPUs in the Scale-Out InfiniBand Network
• Line hint: 773

Context Before

Source: SemiAnalysis AI Networking Model ↗

Evidence

To scale beyond the maximum cluster size of 93,312 GPUs, a 2-plane network can be deployed in which each of the two 800G OSFP cages supporting a GPU are linked to separate leaf switches on different network plans

Context After

2(115,2008002)3=2(1442)3=746,496

② Atomic Claim

若要超過 93,312 顆 GPUs 的 maximum cluster size，可部署 2-plane network，讓支援同一顆 GPU 的兩個 800G OSFP cages 分別連到不同 network plane 的 leaf switches。

• Epistemic Mode: HYPOTHETICAL
• Mapping Status: PARTIAL

③ Semantic Frame

{
  "frame_type": "NARY_RELATION",
  "participants": [
    {
      "node": {
        "id": "04_knowledge_base/GPU",
        "label": "GPUs"
      },
      "role": "supporting_subject"
    },
    {
      "node": {
        "id": "04_knowledge_base/800G",
        "label": "800G"
      },
      "role": "supported_entity"
    },
    {
      "node": {
        "id": "04_knowledge_base/OSFP (Octal Small Form-factor Pluggable)",
        "label": "OSFP"
      },
      "role": "supported_entity"
    }
  ],
  "qualifiers": {
    "condition_text": null,
    "numeric_mentions": [
      "93,312",
      "2",
      "800"
    ],
    "temporal_mentions": []
  },
  "relation_type": "SUPPORTS"
}

④ Canonical Entity Mapping

Role	Surface Label	Canonical Target
supporting_subject	GPUs	GPU
supported_entity	800G	800G
supported_entity	OSFP	04_knowledge_base/OSFP (Octal Small Form-factor Pluggable)

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。