AMD CEO Lisa Su said the company is experiencing an extraordinary period for AI compute following the release of its second-quarter earnings on Wednesday [1].
This surge in demand signals a shift in how businesses use artificial intelligence. As enterprises move from testing technology to deploying it in live production, the need for high-performance hardware increases, creating a critical growth window for chipmakers.
Speaking on CNBC’s “Squawk on the Street,” Su said that while AMD’s Q2 sales topped analysts' estimates [4], the company's shares slipped in pre-market trading [1]. Despite the short-term stock movement, Su said there is tremendous momentum within the company's data-center business [2].
Su said the current market has reached an inflection point [3]. She said that the transition from experimentation to production is driving a need for significantly more processing power. "Customers want a lot more compute," Su said [5].
This demand is fueling aggressive growth projections for the company. According to Su, AMD’s data-center business is on track to grow well over 100% by 2027 [6].
Su said the current environment is unique because of the scale of deployment. "We’re in an extraordinary time for AI compute right now," Su said [1]. The company continues to position its hardware to capture this shift as enterprises scale their AI infrastructure to handle real-world workloads [3].
“"Customers want a lot more compute."”
The transition of AI from the 'experimentation' phase to 'production' means that companies are no longer just testing chatbots or small models, but are integrating AI into core business operations. This requires a massive increase in physical hardware and compute capacity, suggesting that the AI boom is moving from speculative software interest to tangible infrastructure investment.
