AMD Disputes ‘AI Bubble’ Claims, Advocates for $500 Billion AI Market, and Attributes Chip Price Increases to Rising Costs

AMD Disputes ‘AI Bubble’ Claims, Advocates for $500 Billion AI Market, and Attributes Chip Price Increases to Rising Costs

The information provided here does not constitute investment advice. The author does not hold positions in any of the mentioned stocks.

In a recent presentation at a Citigroup conference, AMD’s Chief Financial Officer, Jean Hu, disclosed that the company is not initiating new chip production for its products intended for China, despite having received necessary licenses from the Trump administration. Alongside AMD’s Vice President of Investor Relations, Matthew Ramsay, she articulated the firm’s strategies regarding the expanding market for artificial intelligence (AI) and the increasing costs associated with production, as well as their implications for enhancing profit margins—an essential factor influencing stock performance.

AMD CFO Highlights Early Stages of AI Development without Commenting on Possible Bubble

Hu and Ramsay’s dialogue at the Citigroup event encompassed various critical topics. Similar to NVIDIA, AMD had to write off a portion of its GPU inventory intended for the Chinese market due to complications arising from US regulatory restrictions.

Hu mentioned that AMD is not commencing production of new wafers for the MI308 GPU, which is specifically designed for the Chinese market. The CFO emphasized that the company is focused on depleting its existing inventory. Future investments in China will depend on the firm’s ability to secure licenses for next-generation products, as indicated by Hu.

Ramsay provided insights into the competitive landscape in China, stating, “I think there’s a greater demand for AI processing silicon within China than the capacity to produce that silicon locally.”Due to US sanctions, Chinese semiconductor manufacturers face significant limitations in creating advanced chips. For instance, Huawei, China’s leading technology firm, encounters challenges procuring cutting-edge chips that involve US-origin technology. Ramsay expressed that while AMD is eager to meet the demand for AI chips in China, understanding the immediate landscape has proven challenging amid ongoing fluctuations.

During the discussion, the AMD executives were queried about potential over-ordering of AI chips and the concerns related to the perceived AI bubble. Hu pointed to capital expenditures from hyperscalers in the second quarter, stating that there is “tremendous evidence for AI adoption.”She believes this data indicates not only improved returns on investment across various platforms but also enhanced productivity.

Data Center AI Accelerator TAM graph shows inference growing at >80% CAGR from $45B in 2023 to >$500B in 2028, source: AMD.” title=”Data Center AI Accelerator TAM graph shows inference growing at >80% CAGR from $45B in 2023 to >$500B in 2028, source: AMD.” width=”728″ height=”421″ loading=”lazy” class=”wp-image” src=”https://cdn.thefilibusterblog.com/wp-content/uploads/2025/09/Screenshot-2025-09-04-at-4.15.35-AM-728×421-1.webp”/><figcaption>AMD’s presentation earlier this year detailing a $500 billion Total Addressable Market (TAM) for AI.</figcaption></figure> <p>The AMD CFO also noted that increased AI adoption has positively influenced AMD’s productivity and management of personnel. She shared anecdotal evidence of widespread AI adoption among other companies. In her view, the industry remains in a nascent phase of AI adoption, making it premature to gauge its full impact.</p> <p>Moreover, Hu shifted the focus to AMD’s CPU segment, asserting that the trend in AI adoption is also driving demand for general computing products. She acknowledged the cyclical nature of the industry, asserting that while there are fluctuations over time, AI represents “probably a once-in-a-lifetime opportunity we are witnessing.”</p> <p>Concerning the rise in AI chip prices, Hu explained that as AMD introduces new features with each product launch, the costs of materials invariably increase. This factor contributes to higher retail prices, but AMD’s commitment lies in balancing total ownership costs with gross profit margins for its customers.</p> <p>In a subsequent inquiry, Citi’s Christopher Danley asked AMD executives how they arrived at their $500 billion estimate for the total addressable market for AI. In response, Ramsay mentioned that various factors, including inference use cases and the expansion of customer data sets across industries, significantly drive this outlook.</p> <p>Ramsay articulated, “I’ve often asserted that as time progresses, more capital and operational expenditures across various sectors will gravitate toward high-performance computing, with AI being a substantial catalyst for that change.”</p> <p>He emphasized that AI represents “the most significant transformation in computing since the advent of the Internet, ”suggesting that AMD’s total addressable market could indeed surpass the $500 billion figure previously cited by CEO Lisa Su earlier this year.</p> <p><a class=Source & Images

Leave a Reply

Your email address will not be published. Required fields are marked *