AMD Shifts Focus to AI Chips, Scaling Back Flagship Gaming GPUs

AMD has made it clear: the company is prioritizing AI chips over high-end gaming GPUs. In a recent strategic shift, AMD announced plans to merge its RDNA gaming graphics and CNDA data center efforts into a single universal architecture known as “UDNA,” focusing primarily on artificial intelligence (AI).

In a candid interview with Tom’s Hardware, Jack Huynh, AMD's Senior Vice President for Computing and Graphics, explained the company’s new approach. Huynh revealed that instead of chasing the flagship GPU market currently dominated by Nvidia, AMD’s focus will be on increasing its presence in the broader market by offering more affordable GPUs. This shift is aimed at growing AMD’s market share, which will in turn attract more developers to optimize their software for AMD’s platform.

Huynh outlined AMD’s strategy to achieve 40% market share in gaming graphics, stating, “If I tell developers that I’m only going after 10% of the market, they’ll say, ‘Good luck, but we’ll stick with Nvidia.’ But if I show them a plan for 40%, they’re in.”

While AMD may be pulling back from the ultra-high-end gaming GPU space, the company still aims to be a leader in AI data centers, where it already has significant market share with its EPYC CPUs. Huynh believes AMD can expand its data center business by offering top-tier GPUs specifically tailored for AI applications.

Though the company is refocusing on AI, Huynh reassured gaming enthusiasts: “I love gaming. It remains a strategic pillar for AMD.” He acknowledged that while the company is prioritizing AI for now, it’s not ruling out a return to producing “King of the Hill” GPUs in the future.

AMD’s pivot mirrors rival Nvidia’s similar move toward AI. Like Nvidia, AMD’s financial reports show that data center products now account for a large portion of its revenue, prompting both companies to speed up their development cycles for new architectures. Huynh also hinted at AMD’s ambition to replicate the success of Nvidia’s CUDA platform with its UDNA initiative, aiming for a universal development platform that can handle both AI and gaming.

While this shift could lead to faster innovation and more value-oriented GPUs in the future, it also means that the gaming GPU market may take a backseat for now. Both AMD and Nvidia face strong financial incentives to prioritize data center chips, where profit margins are higher.

Huynh’s comments suggest that AMD is looking to regain its reputation as a leader in affordable GPUs, particularly in the $300–$400 range, which has become highly competitive and where Nvidia’s offerings have faced criticism. For now, AMD’s focus is on delivering value to gamers without aiming to dethrone Nvidia in the premium segment just yet.