New single-chip processor features significantly enhanced GPU to take on Intel / Nvidia in graphics intensive workflows for workstations
AMD has unveiled its highly anticipated ‘Strix Halo’ AMD Ryzen AI Max PRO, a single-chip processor that looks set to deliver a significant leap forward for APUs (Accelerated Processing Units)— CPUs with integrated GPUs.
The new processor, which is designed for mobile workstations and micro desktop workstations, features up to 16 ‘Zen 5’ CPU cores, a powerful ‘RDNA 3.5’ integrated GPU, and an integrated Neural Processing Unit (NPU) for AI rated at 50 TOPS. Despite the computational potential, the power demands of the new chip are relatively low, featuring a configurable Thermal Design Power (cTDP) of 45W – 120W.
At launch, the new processor will be available exclusively from HP in the HP ZBook Ultra G1 A mobile workstation and HP Z2 Mini G1 A desktop workstation (read this article).
The Ryzen AI Max PRO pushes the message that users no longer need to rely on a separate CPU and GPU, providing a compelling alternative to the dominant combination of Intel (CPU) and Nvidia (GPU).
On paper, the AMD Ryzen AI Max PRO represents a major advancement across multiple fronts, particularly in graphics, where performance is reported to rival that of discrete GPUs.
Compared to the AMD Ryzen AI PRO 300 Series laptop APU, which launched last year, the AMD Ryzen AI Max PRO features significantly more RDNA 3.5 graphics compute units (40 vs 16), which should make it applicable to a much wider range of GPU-intensive workflows, such as real-time visualisation and rendering.
Moreover, with a unified memory architecture, up to 96 GB of the platform’s maximum 128 GB of 8000MT/s LPDDR5X memory can be allocated exclusively to the GPU. Typically, such vast quantities of GPU memory are only available in extremely powerful and expensive cloud-based GPUs.
According to AMD, giving the GPU access to this much memory allows the AMD Ryzen AI Max PRO to handle ‘incredibly large, high-precision AI workloads’. The company states the chip can run a 70-billion parameter large language model (LLM) 2.2 times faster than a 24 GB Nvidia GeForce RTX 4090 GPU.
Meanwhile, the NPU is designed to handle lighter AI tasks, such as those offered by Windows Copilot+ and video conferencing features like AutoFrame, background blur, and noise suppression.
AMD is launching four AMD Ryzen AI Max PRO models, including the top-end AMD Ryzen AI Max+ PRO 395 with 16 ‘Zen 5’ cores, 32 threads, up to 5.1 GHz max boost, 80 MB of Cache, 50 peak TOPS and 40 Graphics cores. The full line up can be seen below.
Non ‘PRO’ versions of the chip are also available for consumers. They feature the same silicon but lack certain security and manageability features, as well as Radeon Pro graphics drivers, which are Independent Software Vendor (ISV)-certified for CAD, BIM, and other professional applications.
What DEVELOP3D thinks
The AMD Ryzen AI Max PRO processor is poised to make a significant impact in the workstation sector. On the desktop, AMD has already disrupted the high-end workstation space with its Threadripper Pro processors, severely impacting sales of Intel Xeon workstations. Now, the company aims to bring this success to mobile and micro desktop workstations, with the promise of significantly improved graphics and power efficiency.
The Ryzen AI Max PRO pushes the message that users no longer need to rely on a separate CPU and GPU, providing a compelling alternative to the dominant combination of Intel (CPU) and Nvidia (GPU).
However, AMD faces challenges, particularly in the realm of professional graphics. While the Ryzen AI Max PRO’s RDNA 3.5 GPU should perform well in CAD and core workflows in viz software like Twinmotion and Lumion, it lags behind Nvidia in ray tracing performance. Furthermore, some AEC-focused ISVs rely on Nvidia GPUs to accelerate certain features, and AMD’s graphics drivers could certainly benefit from more optimisation in some professional 3D applications.
In the AI space, although the GPU’s large memory capacity shows great promise, software compatibility will be a key factor in determining the success of the chip, as Nvidia currently holds a commanding lead in AI workloads.
Ultimately, while the AMD Ryzen AI Max PRO shows impressive potential and positions AMD as a strong contender in both mobile and micro desktop workstations, its success will hinge on gaining support from other major workstation OEMs and overcoming challenges related to graphics and software. To truly compete with Intel / Nvidia, AMD will need to match the maturity and ecosystem support that Nvidia currently offers.
