Here’s a bold statement: AMD’s future in the GPU market could hinge on a single decision—whether to bring FSR 4 to older Radeon GPUs. And this is the part most people miss: while FSR 4 has been a game-changer for Radeon, its exclusivity to RDNA 4 graphics cards feels like a missed opportunity. Let’s dive into why this matters and why AMD’s next move could shape its reputation for years to come.
When AMD unveiled FSR 4 last year, it was a watershed moment. Finally, Radeon users had an upscaling solution that rivaled Nvidia’s DLSS in image quality. But there was a catch—FSR 4 was locked to RDNA 4 GPUs, leaving owners of older hardware out in the cold. That is, until a leaked INT8 build of FSR 4 surfaced, proving it could run on RDNA 3 and RDNA 2 GPUs, albeit in an experimental form. But here’s where it gets controversial: AMD has yet to commit to official INT8 support, leaving gamers and enthusiasts scratching their heads.
On RDNA 4 GPUs, FSR 4 leverages hardware-accelerated FP8 math for its AI upscaling, a feature absent in older Radeon architectures. However, the leaked INT8 version demonstrates that, while less optimized, FSR 4 can work on older hardware. Gamers have already tested it, confirming its functionality—though it’s clearly not as polished as the RDNA 4 variant. When Hardware Unboxed pressed AMD for updates on INT8 support, the response was underwhelming: “No updates to share at this time.”
Here’s the kicker: AMD would be making a strategic blunder by not extending FSR 4 to RDNA 3 and RDNA 3.5 GPUs. With new RDNA 3.5 hardware already on the market, failing to support FSR 4 on these products feels like a self-inflicted wound. It’s not just about performance—it’s about goodwill. Nvidia, for instance, took a more inclusive approach with DLSS 4 and 4.5, allowing older RTX cards to use these newer models via the DLSS Override feature. Sure, performance isn’t optimal, but gamers still have the choice. Even RTX 20 series owners can access DLSS 4.5 upscaling. Nvidia gave users the freedom to prioritize speed or quality, and AMD should follow suit.
Bringing FSR 4 to older RDNA GPUs, even if it’s slower than FSR 3.1, would be a win-win. The image quality improvements alone would justify the trade-off, and it would earn AMD significant consumer trust. It would also give RDNA 3.5 laptops and handhelds a much-needed visual boost. So, why hasn’t AMD pulled the trigger? Is it technical limitations, or is there a larger strategy at play? Here’s a thought-provoking question for you: Should AMD prioritize cutting-edge performance on new hardware, or should it focus on supporting its existing user base? Let’s debate this in the comments—your take could shape the conversation.
If you’re as curious (or frustrated) as I am about AMD’s stance, join the discussion on their lack of commitment to FSR 4 on older Radeon GPUs here. The future of FSR 4—and AMD’s reputation—may depend on it.
