clm_6b958e8c58fe4cc1ac90cbc7792ccd7a_r1

VR2-0690

① SA Source

• Source: 開啟完整 SA 文章
• Section: Ethernet-Based Cluster Deployments
• Line hint: 867

Context Before

Source: SemiAnalysis AI Networking Model ↗

Evidence

As the per GPU bandwidth continues to scale to 3.2T, it is not hard to imagine a 4-plane network using SN6800 switches, where each 3.2T GPU is split four ways to connect to four different switch boxes at 800G per link

Context After

In fact, if you have not already noticed – there are strong parallels between the X800-Q3400 switch (non-CPO) explained earlier in the article and a CPO switch in that both facilitate high-radix, multi-plane networking architectures while encasing within the box complex cabling that saves customers on the hassle of cable management.

Outside of the NVIDIA ecosystem, the main switch ASIC players are Broadcom, which will be manufacturing the Tomahawk 6 and Tomahawk 6 CPO ASICs, as well as Cisco, which recently announced the G300 102.4T ASIC. There are two flavors of hyperscaler backend network deployments:

② Atomic Claim

隨每顆 GPU bandwidth 持續提升至 3.2T，可想像使用 SN6800 switches 建立 4-plane network：每顆 3.2T GPU 分成四路，以每 link 800G 連接 4 個不同 switch boxes。

• Epistemic Mode: ASSERTED
• Mapping Status: PARTIAL

③ Semantic Frame

{
  "frame_type": "NARY_RELATION",
  "participants": [
    {
      "node": {
        "id": "04_knowledge_base/GPU",
        "label": "GPU"
      },
      "role": "user_or_subject"
    },
    {
      "node": {
        "id": "04_knowledge_base/SN6800",
        "label": "SN6800"
      },
      "role": "used_entity"
    },
    {
      "node": {
        "id": "04_knowledge_base/800G",
        "label": "800G"
      },
      "role": "used_entity"
    }
  ],
  "qualifiers": {
    "condition_text": null,
    "numeric_mentions": [
      "3.2",
      "4",
      "800"
    ],
    "temporal_mentions": []
  },
  "relation_type": "USES"
}

④ Canonical Entity Mapping

Role	Surface Label	Canonical Target
user_or_subject	GPU	GPU
used_entity	SN6800	SN6800
used_entity	800G	800G

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。