NIEK2-0155

① SA Source

Context Before

Something interesting about the system is the important role the FPGAs play. Nvidia refers to the FPGAs as “Fabric Expansion Logic” which serves multiple purposes. First, they act as a NIC which converts the LPU’s C2C protocol into Ethernet to connect to the Spectrum-X based ethernet scale-out fabric. It is this scale-out fabric through which the LPUs connect to GPUs in the decode system.

Second, the LPUs also traverse through the FPGAs to reach the host CPU, with the FPGAs converting C2C to PCIe to the CPU.

Evidence

This pool of memory can be used for KVCache if the user wants the entire decode process served by the LPX

Context After

On the front panel there are 8 x OSFP cages for cross-rack C2C, while there will be 2 cages (likely QSFP-DD) that goes to the Spectrum-switches that is used to connect the LPUs and the GPUs for the disaggregated decode system. We will share more about this when we describe the network.

LPU Network

② Atomic Claim

若使用者希望整個 decode 流程都由 LPX 執行,這個記憶體池可用來儲存 KVCache

  • Epistemic Mode: HYPOTHETICAL
  • Mapping Status: PARTIAL

③ Semantic Frame

{
  "frame_type": "NARY_RELATION",
  "participants": [
    {
      "node": {
        "id": "04_knowledge_base/Decode",
        "label": "decode"
      },
      "role": "user_or_subject"
    },
    {
      "node": {
        "id": "04_knowledge_base/Groq LPX",
        "label": "LPX"
      },
      "role": "used_entity"
    },
    {
      "node": {
        "id": "04_knowledge_base/KV cache",
        "label": "KVCache"
      },
      "role": "used_entity"
    }
  ],
  "qualifiers": {
    "condition_text": null,
    "numeric_mentions": [],
    "temporal_mentions": []
  },
  "relation_type": "USES"
}

④ Canonical Entity Mapping

RoleSurface LabelCanonical Target
user_or_subjectdecodeDecode
used_entityLPX04_knowledge_base/Groq LPX
used_entityKVCache04_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。