VR2R-0518
① SA Source
- Source: 開啟完整 SA 文章
- Section:
Bi-directional SerDes for Scale-Up - Line hint:
647
Context Before
Let’s step through key features of Rubin’s networking and the architectures that are likely to be built around Rubin scale-up and scale-out networks.
Bi-directional SerDes for Scale-Up
Evidence
The doubling of bandwidth per logical GPU from NVLink 5 in GB300 NVL72 to NVLink 6 in Vera Rubin NVL72 are made possible by using a simultaneous bi-directional SerDes for the copper backplane instead of increasing the modulation or baud rate. Whereas NVLink 5 delivers 224G per electrical lane, NVLink 6.0 delivers 448G per electrical lane. Each electrical lane is one differential pair (DP) consisting of two conductors that carry equal magnitude, and opposite polarity signals.
Context After

Source: Nvidia VR NVL72 Component BoM and Power Budget Model ↗
② Atomic Claim
從 GB300 NVL72 的 NVLink 5 到 Vera Rubin NVL72 的 NVLink 6,每個 logical GPU 的 bandwidth 翻倍;原文認為這是藉由 copper backplane 上 simultaneous bi-directional SerDes 達成,而不是提高 modulation 或 baud rate。
- Epistemic Mode:
ASSERTED - Mapping Status:
COMPLETE
③ Semantic Frame
{
"frame_type": "NARY_RELATION",
"participants": [
{
"node": {
"id": "04_knowledge_base/GPU",
"label": "GPU"
},
"role": "gpu"
},
{
"node": {
"id": "04_knowledge_base/NVLink5",
"label": "NVLink 5"
},
"role": "previous_interconnect"
},
{
"node": {
"id": "04_knowledge_base/GB300",
"label": "GB300"
},
"role": "previous_platform"
},
{
"node": {
"id": "04_knowledge_base/NVLink",
"label": "NVLink 6"
},
"role": "new_interconnect"
},
{
"node": {
"id": "04_knowledge_base/Vera Rubin",
"label": "Vera Rubin"
},
"role": "new_platform"
},
{
"node": {
"id": "04_knowledge_base/SerDes",
"label": "SerDes"
},
"role": "signaling"
},
{
"node": {
"id": "04_knowledge_base/Backplane",
"label": "backplane"
},
"role": "medium"
}
],
"qualifiers": {
"condition_text": "source presents this mechanism hypothetically; instead of increasing modulation or baud rate",
"numeric_mentions": [
"2x"
],
"temporal_mentions": []
},
"relation_type": "BIDIRECTIONAL_SERDES_ENABLES_NVLINK_BANDWIDTH_DOUBLING"
}④ Canonical Entity Mapping
| Role | Surface Label | Canonical Target |
|---|---|---|
| gpu | GPU | GPU |
| previous_interconnect | NVLink 5 | NVLink5 |
| previous_platform | GB300 | GB300 |
| new_interconnect | NVLink 6 | NVLink |
| new_platform | Vera Rubin | 04_knowledge_base/Vera Rubin |
| signaling | SerDes | SerDes |
| medium | backplane | Backplane |
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
從 GB300 NVL72 的 NVLink 5 到 Vera Rubin NVL72 的 NVLink 6,每個 logical GPU 的 bandwidth 翻倍;原文認為這是藉由 copper backplane 上 simultaneous bi-directional SerDes 達成,而不是提高 modulation 或 baud rate。
V2 Semantic Frame Draft
{
"frame_type": "NARY_RELATION",
"participants": [
{
"node": {
"id": "04_knowledge_base/GPU",
"label": "GPU"
},
"role": "gpu"
},
{
"node": {
"id": "04_knowledge_base/NVLink5",
"label": "NVLink 5"
},
"role": "previous_interconnect"
},
{
"node": {
"id": "04_knowledge_base/GB300",
"label": "GB300"
},
"role": "previous_platform"
},
{
"node": {
"id": "04_knowledge_base/NVLink",
"label": "NVLink 6"
},
"role": "new_interconnect"
},
{
"node": {
"id": "04_knowledge_base/Vera Rubin",
"label": "Vera Rubin"
},
"role": "new_platform"
},
{
"node": {
"id": "04_knowledge_base/SerDes",
"label": "SerDes"
},
"role": "signaling"
},
{
"node": {
"id": "04_knowledge_base/Backplane",
"label": "backplane"
},
"role": "medium"
}
],
"qualifiers": {
"condition_text": "source presents this mechanism hypothetically; instead of increasing modulation or baud rate",
"numeric_mentions": [
"2x"
],
"temporal_mentions": []
},
"relation_type": "BIDIRECTIONAL_SERDES_ENABLES_NVLINK_BANDWIDTH_DOUBLING"
}V2 Revision Metadata Draft
{
"epistemic_mode": "ASSERTED",
"mapping_status": "COMPLETE"
}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。