AMD positions AI as a default part of PC and edge computing at CES 2026

• At CES 2026, AMD showed AI becoming a standard part of everyday computing.
• New Ryzen AI processors support local AI workloads on PCs, reducing cloud reliance.

At CES 2026, AMD outlined how it sees artificial intelligence becoming a standard part of personal and commercial computing, rather than a feature limited to high-end systems. The company introduced a broad set of processors and platforms aimed at laptops, desktops, embedded systems, and developers, all designed around on-device AI workloads.

Rather than focusing on a single product line, the announcements point to a wider shift. AI acceleration is now being built into consumer PCs, business laptops, compact desktops, and edge devices at the same time. AMD’s latest releases span mobile processors for Copilot+ PCs, ultra-thin laptops, workstation-class mini-PCs, and embedded systems used in vehicles and industrial settings.

AMD CEO and Chair Lisa Su framed the move as part of a broader inflection point in AI adoption. “Since the launch of ChatGPT a few years ago, I’m sure we all remember the first time we tried it. We’ve gone from one million people using AI to more than one billion active users. That is an incredible ramp. It took the internet decades to reach the same milestone.”

“The PC is being redefined by AIand AMD is leading that transformation,” said Jack Huynh, senior vice president and general manager of AMD’s Computing and Graphics Group. “Across consumer, commercial, and enthusiast systems, we’re delivering platforms that bring high-performance computing, leadership AI, immersive graphics, and a growing software ecosystem that empowers developers and creators, so intelligence is built in, performance and efficiency scale seamlessly, and innovation extends to every form factor.”

Bringing AI acceleration into everyday laptops

One of the core updates is the introduction of the Ryzen AI 400 Series and Ryzen AI PRO 400 Series processors. These chips target consumer and business Copilot+ PCs, with built-in neural processing units designed to handle AI tasks locally rather than relying on the cloud.

Built on AMD’s Zen 5 CPU architecture and second-generation XDNA 2 NPUs, the processors deliver up to 60 TOPS of AI compute from the NPU alone. The Ryzen AI 400 Series is built on the same 4-nanometre process node as the previous generation, with performance gains coming from a combination of higher frequencies, process refinements, and software compile optimizations. AMD said it is already seeing clear improvements on the NPU side, alongside broader platform-level gains.

For business systems, the Ryzen AI PRO 400 Series adds AMD PRO Technologies, including security features, remote management, and long-term platform support. The aim is to allow IT teams to deploy AI-capable laptops across organisations without maintaining separate hardware tiers for different user groups.

From early AI PCs to broader adoption

AMD framed the latest Ryzen AI processors as part of a move from early experimentation toward more routine use. Instead of positioning AI as an optional feature, the hardware is intended to support ongoing tasks such as local inference, content creation, and background automation without constant cloud access.

This approach depends on both hardware and software support. Alongside the processors, AMD announced ROCm 7.2 software support across the Ryzen AI 400 Series, as well as a new AI bundle feature within AMD Software: Adrenalin Edition. The goal is to simplify how developers and users access AI tools on AMD systems, particularly for local model testing and deployment.

Su linked this shift toward local processing to broader limits in global infrastructure.

“The foundation of AI is compute. With this growth in users, we’ve seen a massive surge in demand for global compute infrastructure, growing from about one zettaflop in 2022 to more than 100 zettaflops in 2025.”

High-performance AI in thin and compact systems

AMD also expanded its Ryzen AI Max+ Series with two new processors: the Ryzen AI Max+ 392 and 388. These chips are designed for premium ultra-thin laptops, compact desktops, and small workstations where space and power limits traditionally restrict performance.

According to AMD, the Ryzen AI Max processors are aimed at customers who prioritise GPU capability over raw CPU core counts. As a result, the company expects these processors to be available at more accessible price points than higher-end CPUs with larger core configurations, while still supporting demanding AI, creative, and graphics workloads.

The processors combine Zen 5 CPU cores, Radeon 8060S graphics, and XDNA 2 NPUs within a unified memory architecture. This design allows systems to handle AI models, creative work, and modern games without relying on discrete GPUs.

Gaming performance and product positioning

For desktop users, AMD introduced the Ryzen 7 9850X3D as the successor to the 9800X3D, continuing its focus on cache-heavy designs for gaming workloads.

AMD said the performance uplift between the two processors depends on the title. In games where clock speed plays a larger role, the company is seeing gains of up to 7%. In titles that are less frequency-dependent, the improvement is smaller. AMD plans to provide full generation-over-generation benchmark data to show how performance varies across workloads.

The company also confirmed that the 9800X3D and 9850X3D will coexist rather than one replacing the other, giving users more choice depending on performance needs and price.

A compact platform for AI developers

For developers working directly with AI models, AMD introduced the Ryzen AI Halo developer platform. This AMD-branded mini-PC uses Ryzen AI Max+ processors and is positioned as a local development system for large models and AI workflows.

The system supports up to 128GB of unified memory and can run models with up to 200 billion parameters locally. It also offers up to 60 TFLOPS of RDNA 3.5 graphics performance and supports both Windows and Linux. Out of the box, the platform is optimised for AMD’s ROCm software stack, with pre-installed tools designed to reduce setup time.

Su placed platforms like this within a wider requirement to scale AI across devices, not just data centres.

“To make this possible, AI must exist across every compute platform. We’ll talk about the cloud, where AI runs continuously and delivers intelligence globally. We’ll talk about PCs, where it helps us work smarter and personalise every experience. And we’ll talk about the edge, where AI powers machines that make real-time decisions in the physical world.”

Extending AI to vehicles and industrial systems

Beyond PCs, AMD introduced a new Ryzen AI Embedded processor portfolio aimed at edge deployments in automotive, industrial, and autonomous systems. The P100 and X100 Series processors integrate Zen 5 CPU cores, RDNA 3.5 graphics, and XDNA 2 NPUs into a single chip.

AMD said AI models developed on Ryzen AI Embedded processors can also run on its Versal AI Edge platforms, and vice versa, although portability depends on the specific generation of Versal hardware. Earlier Versal devices differ in quantisation types, memory bandwidth, and CPU characteristics, while second-generation Versal AI Edge processors support newer formats such as MX6 and MX9, enabling broader model support.

“As industries push for more immersive AI experiences and faster on-device intelligence, they need high performance without added system complexity,” said Salil Raje, senior vice president and general manager of AMD Embedded. “The Ryzen AI Embedded portfolio brings CPU, GPU, and NPU capabilities together in a single device, enabling more responsive automotive, industrial, and autonomous systems.”

Video, visualisation, and mixed workloads at the edge

The P100 Series processors are designed to support demanding visual workloads, particularly in automotive and industrial environments.

AMD typically evaluates video performance in terms of maximum resolution, with a common reference point being 4K at 60 frames per second. That workload can be split into multiple streams—such as 10 or 16 concurrent feeds—with some overhead as streams are divided.

According to AMD, testing multi-stream scenarios is a core part of its embedded validation process, with capacity governed by available bandwidth and decoding resources.

A single-portfolio approach to edge AI

AMD argues that its broader portfolio—spanning x86 CPUs and APUs, GPUs, NPUs, adaptive FPGAs, and custom AI silicon—allows it to build single-chip solutions tuned for specific use cases. Rather than relying on multi-chip designs focused only on inference, the company said it optimises the full AI pipeline, from sensing and perception through to control and action, on a single embedded platform.

This approach is aimed at reducing latency and improving efficiency in systems such as robot controllers, digital twins, LiDAR processing, predictive maintenance, and smart cameras, where compute, graphics, and AI workloads need to operate together in real time.

Want to learn more about AI and big data from industry leaders? Check out AI & Big Data Expo taking place in Amsterdam, California, and London. The comprehensive event is part of TechEx and is co-located with other leading technology events, click here for more information.

AI News is powered by TechForge Media. Explore other upcoming enterprise technology events and webinars here.