During CES 2026, Advanced Micro Devices (AMD) made a wide range of announcements across AI, from PC processors to software and data center hardware.
The official press release focuses on client systems and development platforms, but additional Bloomberg reporting shows that AMD is also committed to expanding its AI offerings for enterprise data centers.
Within the client segment, AMD is introducing the Ryzen AI 400 series and the accompanying Ryzen AI PRO 400 line. These processors combine CPU, GPU, and NPU components and are designed for local execution of AI tasks on Windows systems. AMD cites AI computing power of up to 60 TOPS, enabling applications such as image processing, generative AI, and voice functions to run without a cloud connection. The PRO variants are aimed at enterprise environments and include additional management and security features intended for large-scale deployment within organizations.
In addition, AMD is expanding its offering with the Ryzen AI Max+ processors, which target mobile workstations, compact desktops, and notebooks for professional users. These chips combine relatively powerful integrated graphics with AI acceleration and should be suitable for workloads such as content creation and local inference. For developers, AMD also announced the Ryzen AI Halo platform, a compact mini-PC intended as a development and testing environment for AI software on hardware similar to future AI PCs.
On the software front, AMD presented ROCm 7.2, an update to its own open platform for AI development. This version brings more extensive support for Windows and Linux and should simplify the use of AI frameworks on Ryzen systems. The emphasis is on better integration with existing development tools and making training and inference more accessible on consumer hardware.
AI chip for smaller data centers
In addition to these client-focused announcements, Bloomberg reports that AMD also took new steps in the data center market during CES. The company is introducing the MI440X, a new AI chip intended for smaller, enterprise data centers where organizations want to run AI workloads locally and keep data within their own infrastructure. With this, AMD is explicitly targeting companies that do not want to use large-scale cloud platforms.
Bloomberg also reports that during the CES keynote, AMD also focused on the MI455X, the current top model for large-scale AI systems. This chip forms the basis for the Helios system, due to be released later this year and combining a new Venice CPU architecture. During the same presentation, AMD looked ahead to the MI500 series, planned for 2027, which the company claims will deliver a multiple of performance compared to the MI300 generation from 2023.