{
  "$type": "site.standard.document",
  "bskyPostRef": {
    "cid": "bafyreigkdufygmj5grwvdnsplycvkuamqnbo22pc4pcso4cwv266jlwree",
    "uri": "at://did:plc:nfsnsnlxecvkeogkvjmy2u25/app.bsky.feed.post/3mhaflwhewim2"
  },
  "coverImage": {
    "$type": "blob",
    "ref": {
      "$link": "bafkreicx5wpzrn7n4te37xmhrhdma7yuazqwev3m3lz4t6lflgxl5unysi"
    },
    "mimeType": "image/jpeg",
    "size": 51808
  },
  "path": "/article/247094",
  "publishedAt": "2026-03-17T06:15:00.000Z",
  "site": "https://www.cool3c.com",
  "tags": [
    "產業消息",
    "NVIDIA GTC",
    "AI推論",
    "NVIDIA DGX Rubin NVL8",
    "Intel Xeon 6",
    "主機CPU",
    "異構推論"
  ],
  "textContent": "雖然NVIDIA於GTC 2026聚焦自研CPU的Vera處理器，不過仍為主流的x86架構提供NVIDIA DGX Rubin NVL8系統，這套NVIDIA DGX Rubin NVL8則是由Intel的Xeon 6處理器作為Host CPU，Intel強調在AI自大規模訓練轉向及時推論的趨勢，負責管理GPU加速系統之間任務協調、記憶體存取、模型安全與吞吐的Host CPU重要性不斷提升，Xeon 6可提供領先的效能、效率並建立在客戶賴以擴充推論工作的x86軟體生態系。© Intel Corporation▲Xeon 6處理器成為x86架構的DGX Rubin NVL8的Host CPUXe",
  "title": "GTC 2026：Intel Xeon 6成為x86架構之NVIDIA DGX Rubin NVL8的Host CPU"
}