clm_a79b7d9c6f954a21a79149a3f1f04d5c_r1

IX2-0503

① SA Source

• Source: 開啟完整 SA 文章
• Section: Dive into DeepSeek R1 Single Node Results
• Line hint: 710

Context Before

Source: SemiAnalysis InferenceX

Dive into DeepSeek R1 Single Node Results

Evidence

MI355X (SGLang) even beats B200 (SGLang) in throughput performance at lower interactivity levels

Context After

Unfortunately, the year is 2026, and most frontier labs and inference providers are not running FP8 nor single node inference.

This result goes to show that AMDs chips are great and can be extremely competitive with Nvidia if only they could move faster on the software front. Speed is the moat.

② Atomic Claim

在較低 interactivity 水準下，MI355X（SGLang）的吞吐量甚至高於 B200（SGLang）。

• Epistemic Mode: ASSERTED
• Mapping Status: COMPLETE

③ Semantic Frame

{
  "comparison_expression": "在較低 interactivity 水準下，MI355X（SGLang）的吞吐量甚至高於 B200（SGLang）。",
  "entities": [
    {
      "id": "04_knowledge_base/MI355X",
      "label": "MI355X"
    },
    {
      "id": "04_knowledge_base/SGLang",
      "label": "SGLang"
    },
    {
      "id": "04_knowledge_base/NVIDIA B200",
      "label": "B200"
    }
  ],
  "frame_type": "COMPARISON",
  "metric": "INTERACTIVITY",
  "operator": "GREATER_THAN",
  "qualifiers": {
    "condition_text": null,
    "numeric_mentions": [],
    "temporal_mentions": []
  }
}

④ Canonical Entity Mapping

Role	Surface Label	Canonical Target
comparison_entity_0	MI355X	MI355X
comparison_entity_1	SGLang	SGLang
comparison_entity_2	B200	04_knowledge_base/NVIDIA B200

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。