Login to Continue Learning
Hu clarified that AMD is not starting new wafer production for the MI308 GPU and is focused on clearing out its existing inventory. Further investment in Chinese products depends on obtaining licenses for future generations of chips.
Ramsay noted that there is a significant demand for AI processing silicon within China, but local manufacturers are constrained from manufacturing advanced chips due to US sanctions. Huawei, the largest tech firm in China, cannot procure leading-edge chips made with US-origin technology. Ramsay acknowledged the challenge in predicting this market’s short-term dynamics.
When asked about potential over-ordering of AI chips and concerns about an AI bubble, Hu cited second-quarter capital expenditure by hyperscalers as evidence of robust AI adoption. She noted that these firms have improved their return on investment across platforms and increased productivity.
Hu also highlighted the positive impact of AI on AMD’s business, including productivity improvements and personnel management. She believes AI is still in its early stages and too early to gauge its full impact but sees it as a once-in-a-lifetime opportunity.
Regarding rising AI chip prices, Hu explained that each new product includes additional features, increasing the bill of materials and thus selling prices. AMD focuses on total ownership cost for customers while balancing gross margins.
In response to Christopher Danley’s question about factors driving AMD’s $500 billion TAM estimate, Ramsay mentioned inference use cases, customer data set sizes, and their extension into various industries. He stated that AI represents a significant inflection point in high-performance computing, potentially extending AMD’s TAM beyond the $500 billion outlined by CEO Lisa Su.
Ramsay concluded by asserting that AI is “the biggest inflection in computing since the invention of the Internet,” suggesting that AMD’s TAM could extend well beyond the current estimate.
