clm_a29ce8c27ca6488cbdce258de945d0ea_r1

IX2-0252

① SA Source

• Source: 開啟完整 SA 文章
• Section: Nvidia Blackwell Perf TCO Analysis - B100 vs B200 vs GB200NVL72
• Line hint: 418

Context Before

Source: Nvidia GTC 2024 ↗

Evidence

From our testing, Blackwell is so good at large scale MoE inferencing compared to even a strong H100 disagg+wideEP FP8 baseline that it, at 116 toks/s/user, delivers up to 98x better perf on GB200 NVL72 FP4 and up to 100x better perf on GB300 NVL72 FP4!

Context After

Source: SemiAnalysis InferenceX ↗

② Atomic Claim

根據 InferenceX 測試，Blackwell 在大規模 MoE 推論上相較即使是很強的 H100 disagg+wideEP FP8 baseline 仍有巨大優勢；在 116 toks/s/user 下，GB200 NVL72 FP4 最高可達 98 倍，GB300 NVL72 FP4 最高可達 100 倍效能。

• Epistemic Mode: ASSERTED
• Mapping Status: PARTIAL

③ Semantic Frame

{
  "comparison_expression": "根據 InferenceX 測試，Blackwell 在大規模 MoE 推論上相較即使是很強的 H100 disagg+wideEP FP8 baseline 仍有巨大優勢；在 116 toks/s/user 下，GB200 NVL72 FP4 最高可達 98 倍，GB300 NVL72 FP4 最高可達 100 倍效能。",
  "entities": [
    {
      "id": "04_knowledge_base/Blackwell",
      "label": "Blackwell"
    },
    {
      "id": "04_knowledge_base/Mixture of Experts",
      "label": "MoE"
    },
    {
      "id": "04_knowledge_base/H100",
      "label": "H100"
    },
    {
      "id": "04_knowledge_base/FP8",
      "label": "FP8"
    },
    {
      "id": "04_knowledge_base/GB200 NVL72",
      "label": "GB200 NVL72"
    },
    {
      "id": "04_knowledge_base/FP4",
      "label": "FP4"
    },
    {
      "id": "04_knowledge_base/GB300 NVL72",
      "label": "GB300 NVL72"
    }
  ],
  "frame_type": "COMPARISON",
  "metric": "COMPUTE_PERFORMANCE",
  "operator": "MULTIPLE_OF",
  "qualifiers": {
    "condition_text": null,
    "numeric_mentions": [
      "116",
      "98",
      "100"
    ],
    "temporal_mentions": []
  }
}

④ Canonical Entity Mapping

Role	Surface Label	Canonical Target
comparison_entity_0	Blackwell	Blackwell
comparison_entity_1	MoE	04_knowledge_base/Mixture of Experts
comparison_entity_2	H100	H100
comparison_entity_3	FP8	FP8
comparison_entity_4	GB200 NVL72	04_knowledge_base/GB200 NVL72
comparison_entity_5	FP4	FP4
comparison_entity_6	GB300 NVL72	04_knowledge_base/GB300 NVL72

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。