TECH NEWS – AMD (“the reds”) argues that the hype around artificial intelligence isn’t a fleeting phenomenon but a durable trend.

AMD CFO Jean Hu confirmed the company will not start new chip production for its China-focused products despite receiving government licenses. Speaking at a Citigroup conference alongside Matthew Ramsay, AMD’s VP of Investor Relations, Hu discussed the firm’s AI total addressable market, rising costs, and how AMD aims to translate those costs into improved margins in an industry where gross margin drives stock performance.

Like Nvidia, AMD wrote down its China GPU inventory due to U.S. licensing restrictions. Asked whether AMD would alter its China investments or design a modified chip, Hu said the company is assessing whether Chinese customers can buy from the U.S. She added AMD is not starting new wafers for the MI308 GPU (its China-tailored accelerator) and is focused on clearing existing stock. Further investments hinge on securing licenses for the next generation.

Ramsay was frank about China’s market: demand for AI processing silicon exceeds China’s ability to manufacture it. Sanctions block leading-edge production, and Huawei cannot obtain advanced chips made with U.S. technology. While AMD would like to meet that demand, short-term visibility remains limited due to shifting conditions.

On the rumored “AI bubble,” Hu pointed to hyperscalers’ Q2 capex and widespread evidence of AI deployment, which she said boosts both platform returns and productivity—including inside AMD. We are still early in AI adoption, she noted, so it’s too soon to size its ultimate impact.

Crucially, Hu highlighted AMD’s CPU business, saying AI adoption appears to stimulate demand for general compute, too. Cycles have ups and downs, she acknowledged, but AI is “probably a once-in-a-lifetime opportunity.” As for pricier AI chips, every new product adds features and bill-of-materials cost, driving ASPs higher; AMD focuses on improving customers’ total cost of ownership while balancing gross margin.

Citigroup’s Christopher Danley asked what underpins AMD’s roughly $500B AI TAM. Ramsay cited inference use cases, customer dataset size, and cross-industry breadth, calling AI the biggest shift in computing since the internet—implying AMD’s TAM could surpass the $500B CEO Lisa Su outlined earlier this year.

Source: WCCFTech