AMD RDNA 4 GPU Architecture Prioritizes Desktop Graphics Cards, Future Mobile Chips & APUs Potentially On the Horizon

AMD has set its sights on launching its first RDNA 4 graphics architecture products primarily for desktop computers, with potential mobility variants to follow at a later date.

Prioritization of RDNA 4 Architecture for Desktop Graphics Cards

In a recent interview with Notebookcheck, Ben Conard, AMD’s Director of Product Management for Premium Mobile Client Solutions, elaborated on the company’s strategy regarding the next-generation RDNA 4 architecture.

When asked about the potential for RDNA 4 in laptops, Conard emphasized that the initial rollout will focus on desktop platforms. Upcoming offerings will include the newly unveiled Radeon RX 9000 series, encompassing both the RX 9070 and RX 9060 series, aimed at competing with NVIDIA’s RTX 4070/5070 and RTX 4060/5060 series by providing better performance per dollar.

Although AMD has decided to delay the launch of its RX 9000 GPUs, this extra time is intended for optimizing the software stack. There are speculations that this postponement may also facilitate more competitive pricing adjustments against rival products.

VS: Are there future opportunities for RDNA 4 in the laptop market? Currently, there seems to be a lack of AMD discrete GPU laptop models.

Ben: For now, we’re focused on desktop offerings with RDNA 4. I anticipate that we will see these graphics solutions hitting the desktop market first. However, RDNA 4 and subsequent graphics technologies will eventually find their way into mobile platforms, whether through APUs or discrete options.

Despite the current desktop-centric approach, possibilities for mobile launches remain, either through discrete GPUs for laptops or integrated APU solutions. At present, AMD’s RDNA 3.5 architecture is confined to integrated graphics within the Ryzen AI 300 and Ryzen AI Max processors, but upcoming APUs are expected to leverage the more advanced RDNA 4 technology in the near future.

VS: While this may seem ambitious, we’ve heard speculation about a potential merger of RDNA and CDNA architectures.

Ben: Yes, that’s indeed a long-term goal. I find this prospect intriguing, especially with an emphasis on machine learning, which will increasingly drive the client market. Having a unified approach will be significantly beneficial.

Conard also mentioned the emerging UDNA architecture, which aims to integrate both RDNA and CDNA architectures. This unification is poised to reintroduce high-end series graphics cards in AMD’s product lineup, with a strong focus on machine learning and artificial intelligence functionalities—a strategic direction also being pursued by competitors like NVIDIA in their Blackwell series.

For further details, read more at the source.