On Monday, AMD unveiled its latest artificial intelligence chips, aiming to solidify its position as a frontrunner amid competition from industry giants like Nvidia and Intel.
“At AMD, AI is our top priority, and we're entering an incredibly exciting era for the industry as AI revolutionizes virtually every sector, enhances our daily lives, and reshapes the entire computing landscape,” stated Lisa Su, chair and CEO of AMD, during the Computex tech conference in Taipei.
During her keynote address, Su introduced the Ryzen AI 300 series designed for next-generation AI laptops, directly challenging Intel’s upcoming Lunar Lake and Qualcomm’s Snapdragon X. Additionally, in collaboration with Microsoft, these chips will power laptops featuring the tech giant’s AI chatbot, Copilot.
Su also unveiled the Ryzen 9000 series for desktops, labeling them as “the world’s fastest consumer PC processors” tailored for gaming and content creation.
Both chip lines are slated for release in July, following AMD’s April announcement of new processors capable of handling AI workloads—the Ryzen Pro 8040 for laptops and the Ryzen Pro 8000 for desktops.
In the competitive race to develop faster and more potent processors for AI applications, Nvidia recently revealed its next-generation AI chips, dubbed “Rubin,” to succeed the previously announced “Blackwell” model from March.
Nvidia CEO Jensen Huang committed to releasing new AI chip technology annually, a quicker pace compared to the company’s previous biennial timeline. Likewise, AMD intends to introduce new AI chip technology on an annual basis.
AMD also outlined its data center chip roadmap, with the Instinct MI325X accelerators—an enhanced iteration of the MI300 series—set for release in the fourth quarter. The Instinct MI350 series, based on next-generation architecture, is scheduled for launch in 2025, followed by the Instinct MI400 series in 2026.
Su provided a preview of the latest fifth-generation EPYC server processors, slated for release in the latter half of this year, aiming to maintain the leadership position in performance and efficiency established by the AMD EPYC processor family.
Similar to Nvidia, AMD outsources the manufacturing of its chips to foundries, primarily relying on Taiwan Semiconductor Manufacturing Company, the world's largest contract chipmaker.
The Ryzen AI 300, Ryzen 9000, and 5th-gen EPYC chips will be based on the latest “Zen 5” architecture, promising widespread adoption from supercomputers to data centers and PCs, according to Su.
Post a Comment