On December 16 (Note: Per the original information, the actual visit took place on March 17, 2025. The subsequent content is rewritten based on the core information of the original text), Lisa Su, who serves as both the Chairman and CEO of Advanced Micro Devices (AMD), headed a delegation to visit Lenovo Group's headquarters in Beijing. During this visit, she held a meeting with Liu Jun, the Executive Vice President of Lenovo Group and President of its China Region.

This visit came at a significant moment, as Lenovo had just announced a major breakthrough for its inaugural AMD AI large - model training server, the Lenovo WenTian WA7785a G3. This server made history by deploying a 671B (full - version) DeepSeek large model on a single server. It delivered an astonishingly high throughput of 6,708 tokens per second, setting a new benchmark for running large models on a single server.

This remarkable achievement was the result of joint design and collaborative optimization efforts by Lenovo and AMD. Moreover, the two companies have outlined plans to further deepen their cooperation in the realms of AI PCs and servers. Their joint goal is to provide top - notch services to leading cloud computing and AI cluster clients, including industry giants like Microsoft, Meta, and Oracle.