VR2-0235
① SA Source
- Source: 開啟完整 SA 文章
- Section:
BlueField-4 - Line hint:
327
Context Before
KV cache grows linearly with sequence length and multiplicatively with workload parallelism, quickly expanding beyond what any single tier of memory was designed to hold. GPU HBM, while unmatched in bandwidth and latency, is not enough on its own to store KV especially for longer sequence length queries that are becoming popular between turns or tool calls. Host DRAM extends capacity but remains node-bound and limited in aggregate footprint and ultimately has limited capacity. 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.
As we noted in mid-January in our Memory Model note ↗, Nvidia’s ICMS inserts a new G3.5 tier between local SSD (G3) and shared storage (G4), optimized specifically for ephemeral, recomputable KV cache. 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.
Evidence
The challenge with this is that the volumes of SSDs going to ICMS / CMX are quite overblown by the industry. We worked through the math in the Memory model ↗ and Tokenomics model ↗.
Context After
BlueField-4 will be the silicon anchor of this third network. Positioned on the storage array, it terminates NVMe-oF and RDMA traffic at line rate and manages KV movement independently of host CPUs and GPUs. With 2×400G SerDes links providing 800Gb/s of bandwidth, integrated Grace CPU, and LPDDR, BlueField-4 would act as the controller for a distributed context memory fabric. In a preferred DGX-style configuration, a single BlueField-4 per tray may serve four Rubin processors, with the DPU dedicated purely to KV cache traffic and not shared with generic storage I/O.
The new CMX/ICMS ecosystem will likely include leading storage providers such as Weka, DDN, Dell Technologies, NetApp, VAST Data, and others.
② Atomic Claim
SemiAnalysis 已在 Memory model 與 Tokenomics model 中推導相關數學。
- Epistemic Mode:
ASSERTED - Mapping Status:
PARTIAL
③ Semantic Frame
{
"attribute": "UNSPECIFIED_ATTRIBUTE",
"context_nodes": [
{
"id": "04_knowledge_base/Tokenomics Model",
"label": "Tokenomics model"
}
],
"entity": {
"id": "04_knowledge_base/SemiAnalysis Memory Model",
"label": "Memory model"
},
"frame_type": "ATTRIBUTE",
"qualifiers": {
"condition_text": null,
"numeric_mentions": [],
"temporal_mentions": []
},
"value": {
"numeric_mentions": [],
"value_text": "SemiAnalysis 已在 Memory model 與 Tokenomics model 中推導相關數學。"
}
}④ Canonical Entity Mapping
| Role | Surface Label | Canonical Target |
|---|---|---|
| entity | Memory model | 04_knowledge_base/SemiAnalysis Memory Model |
| context_0 | Tokenomics model | 04_knowledge_base/Tokenomics Model |
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。