skip to main content

Lenovo WenTian 联想问天 WA5685 G5 Datasheet

Datasheet

Home
Top
Author
  • Long Ding
Published
20 Aug 2026
Form Number
LP2507
PDF size
2 pages, 2.7 MB

Abstract

联想问天 WA5685 G5 是联想基于 AMD 最新平台打造的一款训推一体 AI 服务器。产品同代兼容 Genoa 和 Turin 两代处理器,支持多种形态的业界主流 PCIe GPU。扩展性上提供了直通和 Switch 互联方案,可为客户的训练、推理场景提供灵活切换方案。能耗设计上兼容风冷和液冷双重散热方案,最低 PUE 可降至 1.25 以下。凭借高算力、高扩展、低能耗优势,产品为中小模型推理、低密度训练提供了最佳选择。

仅限中国大陆使用:联想问天WA5685 G5仅限中国大陆使用

如欲浏览该文件,请按 Download PDF 按钮。

The Lenovo Wentian WA5685 G5 is an integrated training and inference AI server developed by Lenovo based on AMD's latest platform. This product is backward compatible with both the Genoa and Turin processor generations and supports a variety of mainstream PCIe GPU configurations. For scalability, it offers both direct passthrough and Switch interconnect solutions, providing customers with flexible deployment options for both training and inference workloads. In terms of energy efficiency, it supports both air-cooling and liquid-cooling cooling solutions, achieving a minimum PUE below 1.25. With its high computational power, excellent scalability, and low energy consumption, this product is the ideal choice for small-to-medium-scale model inference and low-density training scenarios.

China Mainland only: The Lenovo WenTian WA5685 G5 Server is only available for customers in PRC.

To view the document, please click the Download PDF button.

Related product families

Product families related to this document are the following: