Advanced Micro Devices, Inc. (AMD) is making its boldest move yet to chip away at Nvidia Corp.'s (NVDA) ironclad grip on AI hardware. At CES, the company unveiled a fresh lineup of processors and teased even faster systems coming down the pipeline, all aimed squarely at data centers hungry for AI horsepower.
The Data Center Play: Bigger and Smaller
CEO Lisa Su kicked things off by introducing the MI440X, a new accelerator built for businesses that want to run AI workloads inside their own smaller, on-premise data centers. Think companies that need to keep sensitive data locked down in-house rather than shipping it off to the cloud.
She also showed off the MI455X, AMD's higher-end offering for customers building more powerful AI server systems. According to Bloomberg, Su positioned this as a major upgrade for enterprises scaling up their AI infrastructure.
But AMD isn't stopping there. The company announced its Helios system, built around the MI455X chip, plus a new Venice CPU design set to arrive later this year. Looking further ahead, the next-generation MI500 lineup is slated for 2027. Su made a bold claim: this architecture could deliver up to 1,000 times the performance of the MI300 generation that launched back in 2023.
The Supply Crunch Argument
Su's pitch hinges on a simple thesis: the AI boom has legs because demand for computing power keeps climbing faster than the industry can build capacity. She told Bloomberg that developers and enterprises still can't get enough compute to build what they want, which means the heavy AI spending cycle should continue rolling.
To add some credibility firepower, AMD brought out OpenAI co-founder Greg Brockman, who joined Su on stage to talk up their partnership and hammer home the message that AI workloads need ever-increasing amounts of computing muscle.
AI Escapes the Server Farm
AMD isn't just focused on massive data centers. The company also introduced its new Ryzen AI Embedded processors, designed to bring powerful AI capabilities directly into devices that operate outside traditional server environments.
We're talking cars, factory equipment, medical devices, and robots. These chips are built to help machines process information faster and respond in real time while consuming less power and taking up less physical space.
AMD engineered these processors to handle graphics, computing, and AI tasks all on a single chip. That means companies can build smarter car dashboards, more responsive industrial machinery, and advanced robots without piling on extra hardware. The selling point: add AI features directly into products instead of relying on cloud connectivity.
Consumer Hardware Gets the AI Treatment
On the PC front, AMD rolled out a new wave of chips for laptops and desktops. The Ryzen AI 400 Series targets next-generation Windows AI laptops, while Ryzen AI Max+ chips are aimed at thin-and-light premium notebooks and compact desktops that need extra horsepower for creative workflows and AI tasks.
For the corporate crowd, AMD launched the Ryzen AI PRO 400 Series, built for business laptops with enhanced security features and simpler management tools for IT departments.
The company also announced Ryzen AI Halo, a compact AMD-made system designed to help developers build and test AI applications locally rather than depending entirely on cloud services. AMD paired this with new software updates and bundled tools to make setting up and running AI apps on AMD-powered machines more straightforward.
Gamers Get Their Own Upgrade
AMD didn't forget about gamers. The company unveiled the Ryzen 7 9850X3D, a new desktop processor engineered to deliver better gaming performance than its previous flagship model. AMD also pushed out graphics software updates aimed at making certain games run smoother with sharper visuals.
AMD Price Action: Advanced Micro Devices shares were up 1.07% at $223.45 at the time of publication on Tuesday, according to market data.