AMD Considers Beta for FSR on Older GPUs

AMD's latest generation of AI-powered upscaling technologies, interchangeably referred to as FSR 4 or Redstone, represents a notable leap forward in graphics enhancement. Despite its innovative capabilities, this advanced upscaling solution is not yet officially supported on AMD's older RDNA 3 graphics processing units. However, recent discussions with AMD officials suggest a potential shift in this stance, offering a glimmer of hope for users with previous-generation hardware.

AMD Explores FSR "Redstone" Beta for RDNA 3 GPUs

During a recent interview with PC World, Andrej Zdravkovic, Senior Vice President at AMD and the executive overseeing driver features, including FSR upscaling technologies, addressed the lack of official support for the latest FSR tech on RDNA 3 GPUs. Zdravkovic clarified that the primary hurdle lies in ensuring a consistent performance experience across different hardware generations.

When questioned about the possibility of a "beta version of Redstone that will run on RDNA 3," Zdravkovic's response was cautiously optimistic: "That's currently not in the plan, but thanks for the hint. We may want to think about something like that and provide that to people who want to play with that. The challenge is always how to clearly define that." This statement, while not a definitive commitment, indicates an openness to exploring such an option.

The technical reasons behind this limitation are rooted in the architectural differences between RDNA 3 GPUs, such as the Radeon RX 7800 XT, and the newer RDNA 4 graphics cards, exemplified by the upcoming RX 9070 XT. These differences primarily concern machine learning acceleration capabilities, which impact the performance of AI-enhanced features. While community efforts have demonstrated that the code can be made to run on older RDNA 3 hardware, the results have been inconsistent.

AMD's hesitation stems from a desire to avoid potential support issues arising from a subpar user experience and to uphold its reputation for product quality. Implementing a beta program would allow enthusiastic users to experiment with the technology, even if it doesn't achieve the same peak performance as on newer hardware. This approach could also provide valuable feedback to AMD, potentially enabling them to optimize the technology further for older GPUs.

For now, AMD remains uncommitted to a specific timeline or plan for a beta release. However, the acknowledgment of this possibility offers a hopeful outlook for RDNA 3 users eager to experience the benefits of AMD's cutting-edge upscaling technology.

This discussion highlights the ongoing challenge for hardware manufacturers to balance innovation with backward compatibility. While new technologies often leverage the capabilities of the latest hardware, there's a strong demand from consumers to extend these benefits to previous generations. AMD's willingness to consider a beta program for FSR on RDNA 3 GPUs demonstrates a responsive approach to community feedback and a potential pathway for wider adoption of its advanced upscaling solutions. It underscores the importance of user-driven testing in refining and expanding the reach of emerging technologies, ultimately fostering a more inclusive ecosystem for hardware enthusiasts.