AMD CEO Lisa Su talks about the super cycle and product lineup layout in the AI race.

Nvidia CEO Huang Renxun has been immersed in the high-tech industry for many years. After intense media exposure, Nvidia's stock price skyrocketed in the post-pandemic era, becoming a phenomenon-level stock. Nowadays, Wall Street almost reports Nvidia news every day. Coincidentally, a member of Huang Renxun's family, also from Tainan, is Su Zifeng, who has an impressive academic background and happens to be Huang Renxun's cousin. She entered the AI race ten years ago and has now become Huang Renxun's competitor at AMD.

Su Zifeng is attempting to build a complete AI ecosystem and compete in Full Stack development. She focuses on three core competitive points of AI chips: performance, price, and supply chain.

Recently, Su Zifeng, CEO of AMD Advanced Micro Devices, has changed her low-profile attitude and frequently appeared on TV and news media for interviews. While she used to communicate with many financial journalists online, she rarely revealed her true self in the media. With AMD's stock steadily rising and its product line and supply chain developing steadily,reports predicting an uptrend in AMD's market are increasing, and analysts have listed AMD as a solid investment target.

The AI Race is Just Beginning, Expected to Reach $400 Billion in 2027

During Goldman Sachs' Communacopia investment conference, Su Zifeng, interviewed by Yahoo Finance reporter Brian Sozzi, predicted that the overall AI market will reach $400 billion in 2027. AMD is prepared for the super cycle of the AI race and will launch the MI 325X by the end of the year, aiming to outperform other AI chips in the market.

With nearly 12 years of experience as Vice President of Semiconductor R&D at IBM, Su Zifeng moved to AMD, bringing with her extensive product development experience and solid management skills. She actively invests in acquiring AI startups and is fully committed to developing the Instinct MI 300 series products.

AMD Emphasizes AI Performance, Expected to Surpass Nvidia

Performance is the highlight of AMD. AMD claims that its EPYC processors and Instinct accelerators can provide superior performance for AI data centers compared to Nvidia. The MI 325x is an updated version of the MI 300 accelerator, known as a monster-level memory with larger capacity, offering excellent computing capabilities to handle the complex and crowded workloads of AI data. Following the launch of MI 325x by the end of the year, MI 350 will debut next year, and MI 400 is planned for release in 2026. Nvidia has always refrained from comparing its products to other similar competitive products.

Acquisition of ZT Systems Advances Towards Full-Stack AI Database Development

AMD announced on August 19th the acquisition of ZT Systems for $4.9 billion, moving towards Full Stack development. ZT Systems is a cloud and telecom service provider offering computing, storage, and accelerator solutions.

Supply Chain from the U.S. Aligns with U.S. Priority Policy

When asked by Yahoo Finance reporter about how AMD addresses supply chain issues, Su Zifeng mentioned that AMD has always been a partner of TSMC. However, AMD's chips will come from TSMC's factory in Arizona, USA, rather than Taiwan.While the global competition for AI chips is fierce, geopolitical and economic effects remain the primary considerations. AMD's chip procurement follows the "U.S. priority" principle, ensuring that the supply chain remains uninterrupted by political or other external factors and effectively controls transportation costs.