AMD and Intel Form a Rare Chip Coalition to Win the A.I. Arms Race
In an unexpected yet groundbreaking move, Intel and AMD (AMD) —longtime semiconductor rivals—have joined forces in a bid to strengthen their foothold in the A.I. arms race. At this year's Lenovo Tech World event on Tuesday (Oct. 15) in Seattle, Wash., the two companies announced a coalition called the x86 Ecosystem Advisory Group. The group will act as a collaboration hub to ensure that upcoming AMD and Intel chips are "consistent and compatible" across computing applications and meet the growing demands of generative A.I. (genAI)—whether in data centers, cloud platforms, or consumer devices. While the coalition is centered around AMD and Intel, it will seek insights and feedback from a slew of tech companies, including Broadcom, Dell, Google, Hewlett Packard Enterprise, Lenovo, Meta (META) , Microsoft (MSFT) , Oracle and Red Hat.
Sign Up For Our Daily Newsletter Thank you for signing up! By clicking submit, you agree to ourDespite their fierce competition in the CPU and GPU markets, Intel and AMD's history of working together on key computing standards like PCI, PCIe and USB sets a strong foundation for the latest initiative. x86 is a type of chip architecture that has been the backbone of modern computing for over 40 years. Intel processors like the Pentium and Core i3/i5/i7/i9 and AMD's Ryzen Athlon are some of the most well-known x86-based CPUs.
The new advisory group is intended to set a precedent for expanding the x86 architecture's influence across the tech landscape and enhancing its computing prowess for use cases where industry-wide standardization is crucial, particularly in genAI. Founding members, including Broadcom CEO Hock Tan, termed the collaboration a "crossroads in the history of computing." Through a video message played at the event, Google Cloud CEO Thomas Kurian said that the partnership will unlock "new levels of performance and efficiency."
"The future of A.I. is personalized and on-device," Gelsinger said during his presentation. "Rather than relying on cloud servers for data sharing, the next generation of A.I. will utilize powerful, centralized A.I. processing directly on devices, eliminating the dependency on internet connection while maintaining the same processing capabilities and responsiveness."
After Gelsinger's presentation, AMD CEO Lisa Su stepped up on stage to unveil a new range of AMD EPYC 9005 Series CPUs and Instinct MI325X processors. Su claimed that the new processors will deliver 28 percent faster data processing, ensuring GPUs can handle even the most demanding A.I. workloads. She said Instinct processors have now been fine-tuned for popular genA.I. models like Stable Diffusion 3 and Meta Llama 3, supporting over a million models available through Hugging Face .
"With our 5th Gen AMD EPYC, we are pushing the boundaries of computational and A.I. workload handling capability in the market," Su said onstage. She added that leading A.I. models in the industry, such as OpenAI's GPT and Meta's Llama, are now powered by AMD's latest A.I. accelerators. She also announced AMD's collaboration with Lenovo on making A.I.-powered PCs to speed up the widespread adoption of personalized, hybrid A.I. technologies for consumers.
The recent collaboration between Intel and AMD marks a strategic alignment between the two industry giants, to position them as the top producers of x86-based chips and outpace competitors like Nvidia. Interestingly, it appears Intel is now betting on AMD's recent success to remain at the core of computing innovation for years to come—highlighting a significant shift in the dynamics of the tech world rivalry.
