技术档案

AI计算基础设施AI computing infrastructure

训练推理的芯片与系统

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

纽约Lake Mariner托管设施为液冷AI工作负载专门建设。5
Purpose-built for liquid-cooled AI workloads, the facility is engineered to meet the scale, density and resiliency required for next-generation compute.

Meta已在全球基础设施部署AMD Instinct MI300和MI350系列GPU及数百万EPYC CPU。3
Meta has been a close partner over multiple generations, deploying millions of AMD EPYC CPUs and significant deployments of AMD Instinct MI300 and MI350 series GPUs across their global infrastructure.
Fluidstack为Anthropic建设纽约和得克萨斯州定制数据中心。4
Fluidstack is announcing a $50 billion American computing infrastructure deal with Anthropic, building data centers in New York and Texas, with more sites to come.

Meta于2026年2月已使用英伟达GPU至少十年，新协议扩展了双方技术合作范围。6
The partnership is nothing new, as Meta has been using Nvidia graphics processing units for at least a decade, but the deal marks a significantly broader technology partnership between the two Silicon Valley-based giants.

TeraWulf开发、拥有并运营支持AI和高性能计算的大型数字基础设施。1
TeraWulf develops, owns, and operates large-scale digital infrastructure designed to support AI, high-performance computing (HPC), and other advanced compute workloads.
博通与Meta通过XPU定制加速器平台共同设计MTIA芯片。2
Central to this expanding infrastructure footprint is the rapid deployment of Meta’s MTIA silicon, made possible by deep engineering co-design through Broadcom's foundational XPU (Custom Accelerator) platform.
AMD提供EPYC CPU与Instinct MI300、MI350 GPU，Meta已在全球基础设施中部署这些产品。3
Meta has been a close partner over multiple generations, deploying millions of AMD EPYC CPUs and significant deployments of AMD Instinct MI300 and MI350 series GPUs across their global infrastructure.