AMD has indicated that surging demand for its compute products continues to surpass available supply, with the company preparing to significantly boost production capacity in 2027. The announcement comes as the ongoing AI market boom accelerates sales of both CPUs and GPUs, pushing AMD shares to record highs.
The company’s Chief Executive Officer Lisa Su stated that while AMD has already stepped up supply this year, shortages persist, particularly for memory components. She credited TSMC for helping ramp up capacity and pointed out that memory supply remains tight across the industry.
Efforts to address these challenges include closer collaboration with memory suppliers and customers, ensuring that increases in production capacity align with market needs. AMD’s business model remains unchanged amid these adjustments. Su emphasized Taiwan’s pivotal role in AMD’s operations, confirming increased investment in the region and ongoing exploration of advanced packaging technologies.
During her trip to Taiwan, Su met with key partners including Foxconn and plans further discussions with TSMC to secure additional advanced wafer capacity. She also highlighted the importance of continued coordination with memory manufacturers, referencing an upcoming visit to South Korea.
Su remarked that global AI demand is expected to remain elevated over the next several years, driving long-term supply planning at AMD. She reflected on industry commitments to work with independent auditors to ensure AI systems operate as intended, referencing the recent summit led by U.S. President Donald Trump and technology executives as a step toward heightened industry coordination around AI safety.
Recent market trends have lifted AMD’s profile. The introduction of personal AI agents from companies like Meta and OpenAI has shifted some computing needs from graphics processors to central processors, benefiting both AMD and Intel. In the past month, AMD shares climbed 32%, marking a period of exceptional market performance as new AI applications rely increasingly on CPUs.
Meta’s recently popular AI agent, Muse, utilizes AMD’s CPUs, and the company stated it employs a flexible, CPU-agnostic approach to hardware. OpenAI also draws from multiple CPU suppliers. Meanwhile, leading hyperscalers such as Amazon, Google, Meta, and Microsoft primarily deploy Intel or AMD central processors, while also developing in-house chips using Arm architecture.
Competitor Nvidia introduced its Vera CPU earlier this year, aiming to capture a share of a projected $200 billion CPU market by 2030. However, analysts currently see AMD as having a strong product offering.
AMD’s data center revenue more than doubled in the quarter ending June, reaching $6.7 billion and making up nearly 60% of total sales. The company’s leadership indicated that the shift toward agentic AI workloads has significantly boosted CPU demand and expects this trend to continue.
