VR2R-0228_r1

VR2R-0228

① SA Source

• Source: 開啟完整 SA 文章
• Section: BlueField-4
• Line hint: 198

Context Before

Meanwhile, traditional shared storage—architected for durability rather than latency—has more access time and power overhead, making it unsuitable for participation in the decode loop.

Evidence

Nvidia’s ICMS inserts a new G3.5 tier between local SSD (G3) and shared storage (G4), optimized specifically for ephemeral, recomputable KV cache.

Context After

The ICMS requires a dedicated networking layer designed exclusively for KV traffic. Wherever networking is used in this architecture, it is provisioned as a context memory network—isolated from general data movement and optimized for predictable decode latency.

② Atomic Claim

Nvidia 的 ICMS 在 local SSD（G3）與 shared storage（G4）之間插入新的 G3.5 tier，專門針對 ephemeral、可重新計算的 KV cache 最佳化。

• Epistemic Mode: ASSERTED
• Mapping Status: PARTIAL

③ Semantic Frame

{
  "frame_type": "NARY_RELATION",
  "participants": [
    {
      "node": {
        "id": "02_companies/NVDA",
        "label": "Nvidia"
      },
      "role": "vendor"
    },
    {
      "node": {
        "id": "04_knowledge_base/Inference Context Memory Storage",
        "label": "ICMS"
      },
      "role": "platform"
    }
  ],
  "qualifiers": {
    "condition_text": "local SSD, shared storage and KV cache are plain text in this source claim",
    "numeric_mentions": [
      "G3",
      "G3.5",
      "G4"
    ],
    "temporal_mentions": []
  },
  "relation_type": "ICMS_INSERTS_G3_5_MEMORY_TIER"
}

④ Canonical Entity Mapping

Role	Surface Label	Canonical Target
vendor	Nvidia	NVDA
platform	ICMS	04_knowledge_base/Inference Context Memory Storage

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。