SA260324R-0292_r1

SA260324R-0292

① SA Source

• Source: 開啟完整 SA 文章
• Section: CMX
• Line hint: 372

Context Before

CMX is NVIDIA’s context memory storage platform.

Evidence

CMX addresses a growing bottleneck in modern inference infrastructure: the rapid expansion of KV Cache required to support long-context and agentic workloads.

Context After

KV cache grows linearly with input sequence length and number of users and is the primary tradeoff when it comes to prefill performance (time to first token). At scale, on-device HBM does not have enough capacity. Host DRAM extends beyond HBM capacity with an additional tier of cache, but also hits limits on total amount per node, memory bandwidth, and network bandwidth. Enter NVMe storage for additional KVcache offload.

② Atomic Claim

CMX 用來處理現代 inference infrastructure 日益明顯的 bottleneck：支援 long-context 與 agentic workloads 所需的 KV Cache 快速成長。

• Epistemic Mode: ASSERTED
• Mapping Status: COMPLETE

③ Semantic Frame

{
  "additional_nodes": [],
  "frame_type": "RELATION",
  "object": {
    "id": "04_knowledge_base/KV cache",
    "label": "KV Cache"
  },
  "predicate": "ADDRESSES_BOTTLENECK",
  "qualifiers": {
    "condition_text": "long-context and agentic workloads",
    "numeric_mentions": [],
    "temporal_mentions": []
  },
  "subject": {
    "id": "04_knowledge_base/Inference Context Memory Storage",
    "label": "CMX"
  }
}

④ Canonical Entity Mapping

Role	Surface Label	Canonical Target
subject	CMX	04_knowledge_base/Inference Context Memory Storage
object	KV Cache	04_knowledge_base/KV cache

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。