Ryzen AI Max: 5 reasons not to miss this revolution 💥

AMD has unveiled what one executive described as “the most advanced x86 mobile processor ever created”: the Ryzen AI Max and AI Max+, which boast impressive capabilities for handling graphics and AI workloads. 🚀

AMD is positioning this "Strix Halo" chip as a sort of hybrid for graphics and AI workstations, drawing comparisons to Nvidia's current GeForce 4090 in its ability to run AI language models with up to 70 billion parameters. However, the Ryzen AI Max goes beyond this. 💻✨

It's a graphics-capable APU that's pushing the boundaries of discrete GPUs: in the 3DMark Steel Nomad benchmark, for example, the chip delivers 258 percent better graphics performance than the Intel Core Ultra 9 288V (Arrow Lake) CPU. It also delivers excellent AI performance, both with and without the NPU.

The implication behind this announcement is reminiscent of our report from earlier last year, where we noted that the NPU isn't as crucial to AI capabilities as AMD, Intel, or Qualcomm initially hyped. 👀

The NPU is the most efficient AI core in modern CPUs. But when it comes to pure power, the GPU, especially a discrete GPU, far outperforms both the NPU and the GPU on its own. What AI Max seems to be doing is combining some of the best features of the NPU and GPU, along with a powerful AMD Zen 5 CPU.

AMD will offer the Ryzen AI Max in both consumer and "Pro" versions, which will be sold to businesses. At CES, AMD highlighted three winning designs, including the HP ZBook Ultra G1a and Z2 Mini G1a workstations, as well as the Asus ROG Flow Z13 tablet. 🖥️🏆

“This is something very, very special,” said Rahul Tikoo, senior vice president and general manager of AMD’s client computing division, during a recorded briefing for reporters.

“It’s so unique and powerful. It enables incredible performance and redefines what customers can experience, from the power of workstations and thin laptops to incredibly small and powerful micro desktops. Quite simply, this is the most advanced mobile x86 processor ever created.”

AMD Ryzen AI Max and its features

Interestingly, AMD is offering four of the AI Max chips as commercial products: the Ryzen AI Max+ Pro 395, the Ryzen AI Max Pro 390, the Ryzen AI Max Pro 385, and the Ryzen AI Max Pro 380. However, the first three will be offered in non-Pro, more consumer-friendly versions. 🛒

Ryzen AI Max+ 395: 16 cores/32 threads, 5.1GHz turbo, 80MB cache; 40 graphics cores; 50 TOPS peak, 45-120W cTDP
Ryzen AI Max 390: 12 cores/24 threads, 5.0GHz turbo, 76MB cache; 32 graphics cores; 50 TOPS peak; 45-120W cTDP
Ryzen AI Max 385: 8 cores/16 threads, 5.0GHz, 40MB cache; 32 graphics cores; 50 TOPS peak; 45-120W cTDP

The fourth chip, the Ryzen AI Max Pro, is a 6-core, 12-thread model with 22MB of cache, running at 4.9GHz. All of these will be available in the first or second quarter of 2025, according to AMD. 📅

The interesting thing about the AI Max models is that they don't really differentiate themselves by clock speed. Instead, the number of CPU cores, the number of graphics CUs, and, most importantly, the cache size distinguish the three chips. (It's not entirely clear what AMD means by AI Max+ versus the more generic AI Max.)

What this chip No appears to have is the stacked V-Cache that differentiates AMD's Ryzen X3D chips, but rather a considerable amount of regular L3 cache.

In terms of graphics, AMD has only released two RDNA 3.5 GPUs, the Radeon 880M and 890M, which are part of the “Strix Point” CPU architecture that makes up the Radeon AI 300. These chips have 12 CUs and 16 CUs, respectively, which is far fewer than what the new Ryzen AI Max chips include. 🎨🔍

Alternatively, the Radeon 7000 series includes both the Radeon 7600XT and Radeon 7700XT, which feature 32 CUs and 54 CUs. Both GPUs use the older RDNA 3.0 architecture. (Read our Radeon 7600 review, as well as our Radeon 7700XT review for more details.)

AMD executives stated that the Ryzen AI Max series is designed with a 256GB/s coherent memory interface of unknown width, which can address up to 96GB of graphics memory. However, that's close to the RX7600's 288GB/s memory bandwidth; not as critical for graphics, but still critical for Large Language Models (LLMs) that rely on high memory bandwidth to function. 🌐📊

AMD claims the Ryzen AI Max is the world's first Copilot+ processor capable of running a 70 billion parameter language model approximately 2.2 times faster than the Nvidia RTX 4090 24GB while consuming 87% less power.

AMD is currently positioning Ryzen AI Max for artists, developers, and creators. But with such a powerful integrated GPU, can gamers really be left behind? 🎮💪