CES 2026 has made one thing unmistakably clear: we are moving past the “hype” phase of artificial intelligence and into the era of local execution. Across the keynotes and crowded halls of Las Vegas, the message from AMD, Intel, and Qualcomm was identical: AI shouldn’t just live in a data centre; it needs to live in your pocket, on your desk, and even in the robots walking alongside you.
While each of the “Big Three” chipmakers brought their own flavour of innovation to the stage, they are all sprinting toward a future where AI is faster, more efficient, and deeply embedded into the hardware we use every day.
AMD Bets Big on AI PCs and Performance Gains
AMD Chair and CEO Lisa Su kicked off her CES 2026 keynote with a simple but ambitious tagline: "AI for everyone." For AMD, 2026 is about scaling the “AI PC” from a niche enthusiast product to the standard for every laptop and desktop sold.
The star of the show was the Ryzen AI 400 Series. This next-gen processor family is built on a “triad” of new tech: Zen 5 CPU cores, RDNA 3.5 graphics, and the XDNA 2 neural processing unit (NPU). This NPU is the heavy lifter, delivering up to 60 TOPS (trillion operations per second) of AI performance. That's a notable jump from last year's 50 TOPS and puts AMD firmly in the lead for raw local AI compute.
But AMD isn’t just chasing spec numbers. Su highlighted that the Ryzen AI 400 chips offer a 1.3x boost in multitasking and are 1.7x faster at content creation than the competition. For gamers, the company also unveiled the Ryzen 7 9850X3D, an 8-core beast with second-gen 3D V-Cache that hits a boost clock of 5.6GHz. It's being hailed as the fastest gaming processor in the Ryzen 9000X3D lineup, designed to handle AAA titles and background AI tasks simultaneously without breaking a sweat.
Systems featuring the Ryzen AI 400 and the new gaming chips are expected to hit the market in the first quarter of 2026, with desktop variants following in Q2.
Intel Pushes Panther Lake and the 18A Process
Intel used its CES 2026 spotlight to announce a massive architectural comeback. Senior VP Jim Johnson officially launched the Intel Core Ultra Series 3, better known by its codename Panther Lake. This is a landmark release for Intel, as it is the first compute platform built on the Intel 18A (2nm-class) process – the most advanced semiconductor node ever manufactured in the United States.
Intel is leaning heavily into “Efficiency-First” computing. The Panther Lake chips feature up to 16 cores and a new Xe3 graphics architecture, which Intel claims delivers 77% faster gaming performance than the previous generation. They also introduced XeSS3, a new AI-driven upscaling tool capable of “multi-frame generation”- meaning it can generate three additional frames for every one rendered, making 4K gaming on a laptop much more viable.
On the AI front, Panther Lake delivers 50 NPU TOPS and a staggering 180 total platform TOPS. Intel is also bragging about “all-day” battery life, with some optimised designs reaching up to 27 to 40 hours of video streaming. Beyond consumer laptops, Intel announced that Series 3 is officially certified for “edge” environments. This means the same chips in your laptop are now being used in industrial robotics and smart city infrastructure, supporting complex “vision-language-action” (VLA) models that allow machines to see, reason, and act in the physical world.
Pre-orders for Panther Lake laptops opened on January 6, with global availability set for January 27, 2026.
Qualcomm Expands AI PCs – and Looks Beyond Them
Qualcomm's strategy at CES 2026 was arguably the most diverse. While Intel and AMD focused on the traditional PC, Qualcomm spent its time showing how its mobile-first efficiency is taking over both laptops and the burgeoning world of humanoids.
On the PC side, Qualcomm introduced the Snapdragon X2 Plus. Built on a 3nm process, this chip is designed to bring the premium “Copilot+ PC” experience to more affordable price brackets. Despite being the “Plus” model, it doesn’t hold back on AI – it features the same 80 TOPS Hexagon NPU found in the flagship Elite tier. Qualcomm claims it delivers 35% faster single-core performance than its predecessor while using 43% less power, allowing for multi-day battery life in ultra-portable designs.
The bigger story, however, was Qualcomm's leap into “Physical AI.” The company unveiled the Dragonwing IQ10 Series, a specialised robotics processor powered by an 18-core Oryon CPU. This isn’t just a chip for vacuum cleaners; it's designed to be the “brain” for full-size humanoid robots. Qualcomm is already collaborating with industry leaders like Figure.ai, VinMotion, and Booster to put these chips into robots that can navigate warehouses and interact naturally with humans using on-device language models.
Qualcomm’s message was clear: they aren’t just competing for your laptop; they are building the foundation for the next generation of autonomous machines.
Three Strategies, One Direction
The takeaway from CES 2026 is that the era of “Cloud-only” AI is over. AMD is scaling local performance for the masses, Intel is leveraging its manufacturing might to redefine the architecture of the PC, and Qualcomm is proving that the same efficient AI in your phone can power everything from your next laptop to a humanoid helper.
As these chips start shipping throughout 2026, the competition will shift from “how many TOPS can you hit?” to “what can your device actually do?” With hundreds of new designs in the pipeline, the AI PC is no longer a concept – it's the new baseline for everything we build.
