VR2R-0640
① SA Source
- Source: 開啟完整 SA 文章
- Section:
Ethernet-Based Cluster Deployments - Line hint:
849
Context Before
But first, we will explain why we think multi-plane networking architectures are here to stay as an important preface.
Large-scale cluster deployments where cluster sizes exceed 100k GPUs typically utilize multi-plane network architectures because single-plane network architectures do not have enough logical ports at current switch generations to support larger networks without resorting to a high number of switch layers of 3 or more layers.
Evidence
Recall from above that a Vera Rubin NVL72 cluster built with Q3400-X800 switches at 1.6T logical ports per GPU cannot scale beyond the maximum cluster size of 93,312 GPUs. Even if future switch generations continue to double the maximum possible switching capacity per switch box, the per GPU bandwidth is also expected to double, which means that the effective logical port count in a cluster network is unlikely to change.
Context After
What this means is that deployment of large-scale GPU clusters will continue to require multi-plane network architectures. Do note however that multi-plane networking architectures are not limited by size and we have also seen NVIDIA reference architecture deployments with multi-plane clusters that are significantly below 100k GPUs each.
In multi-plane network architectures using the SN6600 switch instead of the CPO switch, each GPU fans out to multiple switch boxes using fiber shuffles. This requires customers to deploy shuffle boxes, patch panels and unwieldy octopus cables outside the switch box that introduce complexity in installation and maintenance.
② Atomic Claim
即使未來每一代 switch 的單一 switch box 最大 switching capacity 繼續倍增，每顆 GPU 的 bandwidth 也預期會同步倍增，因此 cluster network 的有效 logical port 數量可能不會明顯改變。
- Epistemic Mode:
EXPECTED - Mapping Status:
PARTIAL
③ Semantic Frame
{
"attribute": "FUTURE_BANDWIDTH_SCALING_VERSUS_LOGICAL_PORT_COUNT",
"context_nodes": [],
"entity": {
"id": "04_knowledge_base/GPU",
"label": "GPU"
},
"frame_type": "ATTRIBUTE",
"qualifiers": {
"condition_text": "conditional forecast; switch box and logical port are plain text",
"numeric_mentions": [
"2x"
],
"temporal_mentions": [
"future switch generations"
]
},
"value": {
"numeric_mentions": [
"2x"
],
"value_text": "future switch capacity and per-GPU bandwidth are both expected to double, leaving effective logical-port count roughly unchanged."
}
}④ Canonical Entity Mapping
| Role | Surface Label | Canonical Target |
|---|---|---|
| entity | GPU | GPU |
⑤ V2 Draft（需要修正時直接編輯）
Inline repair
只編輯以下三個 Draft 區塊。完成後將
review_action設為REQUEST_REPAIR，再手動執行 Repair Preview。
執行腳本前，Draft 不會寫入 Claim Store 或 Graph。
V2 Atomic Claim Draft
即使未來每一代 switch 的單一 switch box 最大 switching capacity 繼續倍增，每顆 GPU 的 bandwidth 也預期會同步倍增，因此 cluster network 的有效 logical port 數量可能不會明顯改變。
V2 Semantic Frame Draft
{
"attribute": "FUTURE_BANDWIDTH_SCALING_VERSUS_LOGICAL_PORT_COUNT",
"context_nodes": [],
"entity": {
"id": "04_knowledge_base/GPU",
"label": "GPU"
},
"frame_type": "ATTRIBUTE",
"qualifiers": {
"condition_text": "conditional forecast; switch box and logical port are plain text",
"numeric_mentions": [
"2x"
],
"temporal_mentions": [
"future switch generations"
]
},
"value": {
"numeric_mentions": [
"2x"
],
"value_text": "future switch capacity and per-GPU bandwidth are both expected to double, leaving effective logical-port count roughly unchanged."
}
}V2 Revision Metadata Draft
{
"epistemic_mode": "EXPECTED",
"mapping_status": "PARTIAL"
}⑥ 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。