技术档案
AI基础设施AI infrastructure
训练推理专用算力系统
每条都摘自下方编号的来源原文；只写事实，不写预测或建议。
技术简介
- OCI超级集群由英伟达机架级系统、网络以及吉瓦级供电和冷却组成，用于训练和运行大型AI模型。1
OCI Superclusters Help Organizations Bring Frontier AI to Market Faster at Multi-Gigawatt Scale OCI Superclusters combine NVIDIA NVL72 rack-scale systems, Oracle Acceleron networking with NVIDIA Spectrum-X™ Ethernet networking featuring NVIDIA Spectrum™-6 switches and NVIDIA BlueField®-4 DPUs, and gigawatt-class power and cooling to train and serve the largest AI models.
已有应用
- Meta的Muse使用AMD EPYC处理器，Meta已运行Helios机架。2
AMD said that its Helios racks were running at Meta and its EPYC processors were powering Meta's Muse.
- OpenAI已在Abilene使用配备超过10万颗英伟达Blackwell GPU的OCI超级集群训练和运行GPT-6 Astra。1
In Abilene, OpenAI has trained and served GPT-6 Astra on an OCI Supercluster with more than 100,000 NVIDIA Blackwell GPUs.
主要参与者
- Oracle与英伟达宣布Vera Rubin NVL72和Vera CPU已在OCI全面可用。1
Oracle and NVIDIA also expanded their AI infrastructure collaboration with the general availability of NVIDIA Vera Rubin NVL72 and NVIDIA Vera CPUs on OCI.
相关卡片
- 2026-10-28甲骨文申报向股票和债券投资者筹资 200 亿美元，用于 AI 基础设施建设