AMD is widening its Strix Halo lineup. At CES 2026, the company announced two new Ryzen AI Max+ APUs, aimed at users who want strong AI and graphics performance without stepping up to a discrete GPU.
Strix Halo debuted last year as AMD’s answer to high-end thin-and-light laptops and compact desktops that still need serious compute. The platform combines a high-core-count CPU, a Radeon GPU based on RDNA 3.5, and an XDNA 2 NPU rated at up to 50 TOPS, all sharing a large unified memory pool.
Until now, the lineup topped out with the 16-core Ryzen AI Max+ 395, alongside the 12-core Max 390 and 8-core Max 385. Those chips have already found their way into devices like HP’s ZBook Ultra G1a, Asus’ ROG Flow Z13, and a growing list of mini PCs.
The new additions are the Ryzen AI Max+ 392 and Ryzen AI Max+ 388. Both are based on existing SKUs but receive a notable GPU upgrade. AMD is equipping them with a 40-CU RDNA 3.5 iGPU, which the company says can deliver up to 60 TFLOPs of compute.
On the CPU side, little changes. The Max+ 392 keeps its 12-core, 24-thread layout with boost clocks up to 5GHz, while the Max+ 388 remains an 8-core, 16-thread part. The NPU stays the same as well, with AMD continuing to target AI inference and local model workloads.
AMD also used CES to underline its performance claims. In its own testing, a Strix Halo system reportedly outperformed Nvidia’s DGX Spark in token-per-second-per-dollar AI benchmarks, and showed advantages over Apple’s latest M5-based MacBook Pro in selected AI, multitasking, and gaming scenarios. As usual, independent testing will be needed to validate those numbers.
With support from OEMs like Acer, Asus, HP, Lenovo, Framework, and several mini-PC makers, AMD appears intent on pushing Strix Halo beyond niche use cases and into more mainstream professional and enthusiast systems.
Don’t miss a thing! Join our Telegram community for instant updates and grab our free daily newsletter for the best tech stories!
For more daily updates, please visit our News Section.
Comments