 |
| The AMD Strix Halo gets two new additions this year with 60 TLOPs of GPU compute. |
LAS VEGAS – CES 2026 – Just over a year after unveiling its revolutionary Strix Halo architecture, AMD is doubling down on the AI PC revolution. Today at CES, the chipmaker announced the expansion of its Ryzen AI Max+ lineup with two new processors specifically crafted for AI developers and power users, further intensifying its competition with NVIDIA and Apple in the high-stakes arena of on-device artificial intelligence.
Recapping the Strix Halo Disruption
Last year’s introduction of the Ryzen AI Max+ Strix Halo platform was a watershed moment. AMD’s goal was audacious: pack workstation-level performance for AI and content creation into sleek, portable laptops. The secret sauce was a revolutionary design combining a powerful Zen 5 CPU core, a dedicated XDNA 2 Neural Processing Unit (NPU), and a massively integrated Radeon 800M Series GPU—all sharing a pool of ultra-fast unified memory. This "all-in-one" approach eliminated the need for a bulky, power-hungry discrete graphics card without sacrificing performance.
The initial trio—the Ryzen AI Max+ 395 (16-core), 390 (12-core), and 385 (8-core)—quickly found homes in premium devices like the HP ZBook Ultra G1a laptop, the Asus ROG Flow Z13 2-in-1, and a wave of potent mini-PCs from brands like Minisforum and GEEKOM.
Meet the New Contenders: Ryzen AI Max+ 392 & 388
For 2026, AMD is filling out the lineup with two new SKUs designed to offer more granular choices for developers and creators.
- Ryzen AI Max+ 392: Positioned as a supercharged version of the Ryzen AI Max 390, it retains the 12-core/24-thread CPU configuration and 5 GHz boost clock but receives a significant graphics upgrade. The integrated Radeon GPU now boasts 40 RDNA 3.5 Compute Units (CUs), up from the previous generation, delivering a staggering up to 60 TFLOPs of GPU compute power.
- Ryzen AI Max+ 388: Similarly, this chip builds upon the Ryzen AI Max 385. It keeps the 8-core/16-thread design but is now equipped with the same formidable 40 CU GPU.
Both new chips earn the coveted "Max+" branding, a designation AMD reserves for APUs featuring this top-tier 40 CU graphics configuration. The upgraded GPU isn't just for gaming; it provides immense parallel processing power crucial for AI model training, inference, and GPU-accelerated creative tasks.
Performance Claims: Taking on NVIDIA and Apple
AMD didn’t just announce specs—it came armed with competitive performance data. In a direct challenge to NVIDIA’s AI-focused systems, AMD showcased benchmarks for large language model (LLM) inference.
The company claims a system like the $2,500 HP Z2 Mini G1a, equipped with a Ryzen AI Max+ 395 and 128GB of unified memory, delivers significantly better value. In LM Studio running the GPT-OSS 20B model, AMD says it achieves 1.5x higher tokens per second per dollar compared to a $4,000 NVIDIA DGX Spark system (also with 128GB memory). For the larger GPT-OSS 120B model, the advantage widens to 1.7x.
The battle with Apple is equally fierce. Pitting an Asus ROG Flow Z13 (Ryzen AI Max+ 395 at 55W "Silent" profile) against a MacBook Pro 14 with a 10-core M5 chip ("Balanced" profile), AMD claims substantial leads:
- 1.4x faster AI performance
- 1.8x faster multitasking & content creation
- 1.6x faster gaming
These claims highlight AMD’s strategy: offering a single, versatile platform that excels at professional AI workflows, creative applications, and gaming.
Ecosystem Momentum: From Laptops to Handhelds
A chip is only as good as the devices it powers, and here, AMD shows strong momentum. A growing list of OEMs is integrating Ryzen AI Max+ processors into new designs.
Major laptop brands like Acer, Asus, Framework, HP, and Lenovo have next-generation systems in the pipeline. Intriguingly, the platform's power efficiency and graphics muscle have also attracted gaming handheld manufacturers. Devices like the upcoming GPD Win 5 are leveraging Strix Halo to run both Windows 11 and SteamOS, promising a new class of ultra-portable powerhouses.
For desktop enthusiasts and developers, the mini-PC market is exploding with options. Brands like GEEKOM, Minisforum, GMKtec, Beelink, and even newcomers like Corsair and COLORFUL are launching compact systems. These tiny towers offer a compelling, space-saving alternative for AI development stations and high-end home theaters.
Speaking of mini-PCs, one of the most compelling ways to experience this performance today is through systems like the GMKtec EVO-X2. This compact powerhouse leverages the flagship Ryzen AI Max+ 395 APU, offering a desktop-grade AI and creative workstation in a form factor smaller than a textbook.
Buy the GMKtec EVO-X2 mini PC with Ryzen AI Max+ 395 on Amazon
The Bottom Line
With the expansion of the Ryzen AI Max+ family, AMD is systematically addressing the core demand of the AI PC era: scalable, accessible, and uncompromising performance. By targeting the developer community with more options and boldly challenging both NVIDIA's value proposition and Apple's performance hegemony, AMD is ensuring the Strix Halo vision has lasting impact. For professionals, creators, and tech enthusiasts, 2026 is shaping up to be the year where true on-device AI capability becomes mainstream, and AMD is fiercely contesting for the lead.
