AMD CEO Lisa Su stated on Tuesday that the company’s total revenue growth is anticipated to rise to approximately 35% annually over the next three to five years, fueled by “insatiable” demand for AI chips.
Su noted that a significant portion of that would be garnered by the company’s AI data center sector, which it projects to grow by around 80% annually during the same time frame, poised to reach sales in the tens of billions by 2027.
“This is our potential based on the customer engagement we’ve seen, both from announced clients and those collaborating closely with us,” Su informed analysts.
In the end, Su remarked that AMD could attain a “double-digit” market share in the data center AI chip sector within the next three to five years.
AMD shares dropped 3% in after-hours trading. Shares had risen after AMD reported that their gross margins for the coming years are expected to fluctuate between 55% to 58%, surpassing analysts’ predictions.
The AI chip sector is currently led by Nvidia, which dominates with over 90% of the market share, as per various estimates, contributing to a market capitalization exceeding $4.6 trillion, compared to AMD’s approximate $387 billion valuation.
AMD is hosting its inaugural financial analyst day since 2022, as the company finds itself central to a surge in data center investment for AI.
As firms allocate hundreds of billions of dollars in total on graphics processing unit (GPU) chips for developing and powering AI applications like OpenAI’s ChatGPT, they are also seeking alternatives to boost capacity and manage expenses. AMD is the second prominent GPU developer after Nvidia.
In October, AMD declared a partnership with OpenAI where it would sell the AI firm billions in its Instinct AI chips over several years, starting with a sufficient quantity of chips in 2026 to utilize 1 gigawatt of power.
As part of the agreement, OpenAI may acquire a 10% stake in the chip manufacturer. Su also emphasized long-term agreements with Oracle and Meta on Tuesday.
AMD shares have nearly increased twofold so far in 2025.
OpenAI is also assisting AMD in establishing its next-generation systems centered around its Instinct MI400X AI chips, which are set to ship next year.
AMD has communicated that its chips will be capable of being configured into a “rack-scale” system where 72 of its chips collaborate as a single unit, which is crucial for implementing the largest AI models.
If AMD succeeds with its rack, it will catch up with Nvidia’s AI chips, which have been available in rack-scale systems for three product generations.
Su mentioned that the total market for AI data center components and systems is now projected to reach $1 trillion annually by 2030, implying a growth rate of 40% each year. AMD recorded $5 billion in AI chip sales for its fiscal 2024.
This marks an increase from AMD’s earlier prediction of a $500 billion market for AI chips by 2028. However, the revised figure includes central processors (CPU), a crucial type of chip that operates at the core of a computer but isn’t exclusively an AI accelerator like the GPUs produced by Nvidia and AMD.
AMD’s Epyc CPUs remain the company’s most significant product in terms of revenue. It primarily competes against Intel and some smaller Arm-based processors within the CPU arena. AMD also fabricates chips for gaming consoles, networking components, and other devices.
On Tuesday, although AMD placed significant emphasis on its expanding AI sector, it conveyed to shareholders that its legacy businesses are also seeing growth.
“The additional point we want to convey today is all other segments of our business are performing exceptionally well, and that’s a very favorable situation to be in,” Su stated.
