Nvidia 发布面向 agentic AI 的 Vera CPU Rack 规格
Nvidia Unveils Vera CPU Rack Specification for Agentic AI < Semiconductor < 기사본문 - The Elec Inc.
Nvidia 于 10 月 12 日发布 Vera CPU Rack 规格，首次推出以 CPU 为核心的服务器机架设计，面向 agentic AI 负载。按其架构，GPU 版 Vera Rubin NVL72 机架负责传统 LLM 训练与推理，Vera CPU Rack 则处理 agentic AI 负载。该机架基于第三代 MGX 架构，并将 Vera Rubin 机架及 MGX 1.2 的抽屉式计算托盘设计提交至 Open Compute Project 的 Open DC Initiative。MGX 1.2 兼容 19 英寸机架设备，单机架支持最高 250kW，并引入机架级液冷。
推荐理由
AI 评分 60
Nvidia 首次推出以 CPU 为核心的机架规格，读者可了解 agentic AI 负载对数据中心硬件形态的具体影响。
主题
正文
当前显示原文，中文全文译文尚未就绪。
Company submits Vera Rubin MGX 1.2 designs to Open Compute Project
JEONG IL JOO; Published 2026.10.14 18:23; Reply 0
Nvidia unveiled its Vera CPU Rack specification on Oct. 12, introducing a server rack design centered on central processing units (CPUs) for the first time. The specification is optimized for agentic artificial intelligence (AI) workloads.
According to Nvidia, agentic AI relies heavily on CPU computing. Under its architecture, the graphics processing unit (GPU)-based Vera Rubin NVL72 rack handles traditional large language model (LLM) training and inference, while the Vera CPU Rack is designed to process agentic AI workloads.
The Vera CPU Rack is based on Nvidia’s third-generation MGX architecture. MGX is a modular reference specification for data center servers that Nvidia provides to server manufacturers. Partners including Dell, Supermicro and Quanta use the MGX architecture to design and manufacture their own server products.
The Vera CPU Rack integrates computing, networking, power and cooling systems into a common modular platform. Server manufacturers can share the same infrastructure resources while designing dedicated servers for specific purposes.
Nvidia also said it submitted designs for the Vera Rubin rack and drawer-style compute trays based on MGX 1.2, an updated version of its third-generation MGX specification, to the Open Compute Project’s (OCP) Open DC Initiative. OCP is an industry organization that shares designs submitted by its members as common specifications.
The company said the initiative is intended to help server manufacturers develop next-generation AI factory infrastructure based on Vera Rubin and bring it to market more quickly.
MGX 1.2 is compatible with 19-inch server rack equipment and supports up to 250 kilowatts (kW) per rack. It also introduces rack-level liquid cooling. Nvidia said standardized interconnects, reconfigurable input/output (I/O) slots and fanless compute trays reduce assembly and maintenance time.