clm_72af0bc2926c47ea8db400502e1fa22d_r1

GB2-0025

① SA Source

• Source: 開啟完整 SA 文章
• Section: Networking
• Line hint: 263

Context Before

As a quick refresher, the frontend networking is just your normal ethernet network that you use to connect to the internet, SLURM/Kubernetes, networked storage, data loading, model checkpoints. This network is typically 25-50Gb/s per GPU, so on a HGX H100 server, it will be 200-400Gb/s per server, while on a GB200 computer tray node, it will be 200-800Gb/s per server depending on the configuration.

Your backend network is used to scale out GPU-GPU communications across hundred to thousands of racks. This network could either be Nvidia’s Infiniband or Nvidia Spectrum-X Ethernet or Broadcom Ethernet. With the options from Nvidia being way more expensive compared to the Broadcom Ethernet solutions ↗.

Evidence

On Hopper, this network connected 8 GPUs together at 450GB/s each while on Blackwell NVL72, it will connect 72 GPUs together at 900GB/ each.

Context After

Lastly, there is your out of band management network which is used for re-imaging your operating system, monitor node health such as fan speed, temperatures, power draw, etc. Your baseboard management controller (BMC) on servers, PDUs, switches, CDUs are usually connected to this network to monitor and control these IT equipment.

NVLink Scale Up Interconnect

② Atomic Claim

Blackwell NVL72 的 scale-up accelerator interconnect 連接 72 顆 GPU、每顆約 900GB/s；Hopper 則是 8 顆 GPU、每顆 450GB/s。

• Epistemic Mode: ASSERTED
• Mapping Status: COMPLETE

③ Semantic Frame

{
  "comparison_expression": "Hopper: 8 GPUs at 450GB/s each; Blackwell NVL72: 72 GPUs at 900GB/s each",
  "entities": [
    {
      "id": "04_knowledge_base/Hopper",
      "label": "Hopper"
    },
    {
      "id": "04_knowledge_base/Blackwell",
      "label": "Blackwell"
    },
    {
      "id": "04_knowledge_base/GPU",
      "label": "GPU"
    }
  ],
  "frame_type": "COMPARISON",
  "metric": "scale_up_world_size_and_bandwidth",
  "operator": "GENERATION_INCREASE",
  "qualifiers": {
    "condition_text": null,
    "numeric_mentions": [
      "8",
      "450GB/s",
      "72",
      "900GB/s"
    ],
    "temporal_mentions": []
  }
}

④ Canonical Entity Mapping

Role	Surface Label	Canonical Target
comparison_entity_0	Hopper	Hopper
comparison_entity_1	Blackwell	Blackwell
comparison_entity_2	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。