With the HP Amplify Partner Conference coming up this week, it’s a good time to preview the HP ZBook Ultra G1a and the AMD Ryzen™ AI Max PRO Series processors this new 14” mobile workstation uses. Formally announced at CES 2025, the AMD Ryzen™ AI Max PRO Series processors (codenamed "Strix Halo"), was one of the breakout stars of the show -- and for good reason.
As the name implies, the AMD Ryzen™ AI Max PRO Series processor sits at the top of AMD’s mobile product stack. These processors push the envelope of integration and scalability by combining AMD’s latest CPU and NPU technology alongside a powerful integrated GPU with discrete-level performance. The AMD Ryzen™AI Max PRO 395+ is both the first 16-core Ryzen PRO processor AMD has launched, and the first Ryzen processor to ever feature a unified memory architecture that supports up to 128GB of RAM. Systems based on AMD Ryzen™ AI Max PRO Series processors are designed for creative professionals and software developers facing the challenge of adapting, understanding, and prototyping new AI solutions.
Technical Specifications:
At its core, the AMD Ryzen™ AI Max PRO Series is a leap forward for the integration of functionality that is traditionally implemented in multiple discrete components, namely the CPU and the graphics processor (GPU). Its performance, power consumption, and features illustrate the continuing benefits of incorporating new functionality into a single microprocessor.
AMD Ryzen™ AI Max PRO Series processors offer up to 16 desktop-class "Zen 5" CPU cores, a neural processing unit (NPU) based on the AMD XDNA™ 2 architecture, and an integrated AMD RDNA™ 3.5 architecture-based GPU with performance previously only available with solutions that use a discrete graphics card. The official AMD Ryzen™ AI Max PRO Series processor whitepaper covers the chip’s technical details, but I'll touch on a few points here as well.
Historically, the CPU and GPU had their own, separate memory buses, as shown above. This bifurcated arrangement increases the memory bandwidth available to the graphics card but decreases the total amount of memory it can address.
AMD Ryzen™ AI Max PRO Series processors integrate their CPU, GPU, and also the NPU into the same physical die and double up on memory bandwidth by using a 256-bit LPDDR5X memory interface. Additionally, the CPU includes up to 32MB of GPU MALL (Memory Access at Last Level) cache, aka AMD Infinity Cache™ technology, designed to further improve performance. This unified memory architecture gives the top-end AMD Radeon 8060S access to up to 96GB of VRAM out of 128GB total -- vastly more than any other mobile GPU.
This massive pool of memory and its associated bandwidth has real, practical uses. Provisioning a system with this much RAM via a unified memory architecture means enterprises can run complex AI inference tasks or large data visualization models on-device without the usual bottlenecks imposed by limited GPU memory. Whether it’s real-time analytics for finance or high-fidelity simulations for engineering, the AMD Ryzen™ AI Max PRO Series can maintain better data locality to keep performance high.
But the AMD Ryzen™ AI Max PRO Series isn't just a list of features, impressive though that list might be. It's also an example of how AMD is keeping long-term trends in CPU development on track for the entire industry, even as new workloads and use-cases challenge traditional computing architectures.
Putting AMD Ryzen AI Max PRO Series Processors in Historical Context:
Earlier, I referred to the AMD Ryzen™ AI Max PRO Series as a leap forward for integration. The concept of integration has been fundamental to the evolution of CPUs since their invention, but the technical question of what is being integrated has changed a great deal.
Decades ago, integration meant combining all the functional blocks of the processor on the same piece of physical silicon. Once that goal had been achieved, CPUs began assimilating capabilities that were previously handled by external caches and co-processors separately attached to the motherboard. Incorporating these processing elements into the CPU allowed them to benefit from generational manufacturing improvements and resulted in faster performance, better power efficiency, and lower communication latency across the chip.
The long-term impact of this integrative improvement cannot be measured in speeds, feeds, or core counts. To properly capture its scope, one must speak instead of the industries that were created and transformed as CPUs advanced and absorbed additional functionality.
In the 1980s, software developers took advantage of CPU improvements to build the spreadsheets, word processors, and desktop publishing suites that became vanguards of the modern office. Non-linear video editing, 3D rendering, and digital audio workstations revolutionized creativity in the 1990s, as PCs got less expensive and businesses moved online. Even large-scale trends, like the growth of the Internet and its attendant impact on virtually every aspect of modern life, were enabled by a virtuous cycle of silicon engineering advances – and driven by designers like AMD.
Core counts and clock speeds rose throughout the 2000s and 2010s, boosting performance and dramatically increasing the number of CPU cores available per socket. In response, enterprises deployed virtualization and adopted cloud computing to improve server utilization, shrink the number of physical servers they needed to deploy, and reduce costs. Office apps came more collaborative, encouraging coordination across far-flung environments. The PC ecosystem shifted away from tower desktops and towards laptops and compact systems as mobile core counts and battery life increased simultaneously.
The cycle is clear. Every decade, CPU designers integrate 2-3 new, significant capabilities. Every decade, software engineers take advantage of these features to enhance existing products and invent new lines of business. In some cases, advances in CPU integration have been large enough to help spur the creation of entirely new services, markets, and social spaces.
Most recently, CPUs have integrated high-performance AI engines to accelerate new artificial intelligence workloads. AMD was the first manufacturer to bring x86 AI PC processors to market back in 2023, and the AMD Ryzen™ AI Max PRO Series builds on this forward-looking approach by combining a second-generation NPU with the largest, most capable integrated GPU ever featured in an x86 PC.
Viewed in this context, the AMD Ryzen™ AI Max PRO Series takes on added significance. Like its forebears, it integrates new compute capabilities to enable innovative, previously unknown use cases. Critically, these processors are arriving in-market as developers and enterprises grapple with the increased memory and computing requirements of AI workloads. With its discrete-class integrated GPU and the same AMD XDNA 2-based NPU as other AMD Ryzen™ AI 300 Series processor-based systems, the AMD Ryzen™ AI Max PRO Series provides the advantage of two dedicated AI processing solutions depending on what the developer’s needs are. That’s because AMD Ryzen™ AI Max PRO Series processors aren’t just designed to run the applications and workloads of the future – they’re designed to build them.
AI and the Enterprise Market:
An overwhelming majority of corporations are now deploying artificial intelligence or experimenting with it at some level. Microsoft, similarly, is all-in on AI as a distinguishing feature for both commercial and client systems. The OS developer debuted its "Copilot+ PC" initiative in early 2024, alongside a new, dedicated Copilot button on AI PC keyboards. Copilot+ PCs offer advanced AI functionality running directly on-device, but only devices capable of >40 TOPS qualify.
AI PCs are expected to proliferate rapidly in market thanks to the imminent Windows 11 refresh cycle, and broad OEM availability, as discussed by corporate VP Gaston Sandoval in his recent blog. Functional AI deployments and the popularity of hybrid AI (AI that runs both on-device and in the cloud) will rise as companies experiment with the technology.
Devices powered by AMD Ryzen™ AI Max PRO Series processors sit at the top of the AI PC stack. They share the AMD Ryzen AI PRO 300 Series’ AMD XDNA 2 NPU architecture and similarly qualify as Microsoft Copilot+ PCs, but offer substantially more memory bandwidth and a massive pool of configurable VRAM. This allows these systems to support AI inference models and conventional workstation workloads most mobile video cards can't handle.
The AMD Ryzen™ AI Max PRO Series is for engineers, product designers, architects and creators who want professional, ISV-certified graphics without the hit to battery life that typically comes with using a discrete GPU. It's for enterprise customers who need the security, manageability, and reliability features AMD PRO Technologies provide. It's for programmers developing and fine-tuning AI. But most of all, it's for the visionaries who turn technical improvements into boundary-breaking software, from WordStar and Lotus 1-2-3 to the creative applications, operating systems, and AI models that define the leading edge of computing today.
Readers curious for more details on AMD Ryzen™ AI Max PRO Series processors can check the whitepaper, evaluate competitive AI performance via this discussion, or this video discussing the AMD Ryzen AI Max Series in a creative context. The chip has also been discussed alongside the broader evolution of AMD’s integrated GPU strategy by AEC Magazine, and additional information on AMD Ryzen™ PRO mobile workstations, including AMD Ryzen™ AI Max processors, is available here.