When Nvidia announced that it’d be letting older GPUs gain access to DLSS 4, all eyes were on AMD to do the same. Previously, both GPU makers locked their most precious AI-driven frame generation and upscaling tech behind a paywall of sorts: you had to own a GPU from the latest generation in order to try them out.
It didn’t take long for AMD to follow suit, and follow suit it did, with an announcement that just gave older graphics cards a whole new lease on life. With GPU prices through the roof, I dare say this announcement couldn’t have come at a better time.
AMD is bringing FSR 4.1 to older Radeon cards
It’s not just RDNA 4 anymore
In a recent announcement, AMD’s Jack Huynh revealed that the chipmaker would be bringing its latest FSR 4.1 Upscaling to older RDNA GPUs. Many fans were awaiting this announcement, especially since it was revealed that Nvidia did something along those lines with DLSS 4. More than that, a modder was able to get FSR 4 Frame Generation to work on RDNA 3 last year, so users knew it was technically possible. (It was just a matter of “when,” or so we all hoped.)
Until now, FSR 4 was one of the big reasons to look at AMD’s RDNA 4 graphics cards. It marked AMD’s shift to ML-powered upscaling, which meant better image stability, sharper details, and fewer artifacts.
Now, with FSR 4.1 coming to RDNA 3 (RX 7000) in July, and even RDNA 2 in Q1 of next year, gamers who are still running older hardware have a lot to look forward to. A free update that genuinely boosts performance? Yes please. Better yet, AMD says it’ll work across a suite of over 300 supported games at launch, so this isn’t some kind of proof-of-concept. It should have a real impact for many users.
Some questions still remain unanswered, such as how much of the larger FSR suite will be coming to those older GPUs. Unlike AMD’s latest cards, RDNA 3 and 2 cards don’t have AMD’s second-gen FP8 AI accelerators. Without the right hardware, we might not see the full Redstone stack in play, but even just FSR 4.1 upscaling is a huge win.
QuizGreatest GPUs of all time
Trivia challenge
From the 3dfx Voodoo era to modern powerhouses — how well do you know the GPUs that shaped PC gaming history?
GPU HistoryPerformanceMilestonesHardwareBrands
The 3dfx Voodoo2 was a landmark GPU of the late 1990s. What made the Voodoo2 particularly unique compared to most graphics cards of its era?
Correct! The Voodoo2 pioneered a technology 3dfx called SLI (Scan Line Interleave), which allowed two cards to work together by rendering alternating scan lines. This gave enthusiasts a meaningful performance boost and made multi-GPU setups a real consumer option for the first time.
Not quite. The standout feature of the Voodoo2 was its support for SLI (Scan Line Interleave), letting two cards work in tandem. This was a first for consumer graphics and made it the go-to card for serious PC gamers in 1998.
NVIDIA marketed the GeForce 256, released in 1999, with a bold and historic claim. What was that claim?
Correct! NVIDIA coined the term ‘GPU’ specifically to market the GeForce 256, touting its ability to perform hardware Transform & Lighting (T&L) directly on the graphics chip. Before this, T&L calculations were handled by the CPU, so this was a genuine architectural leap.
Not quite. NVIDIA invented the term ‘GPU’ for the GeForce 256, which was the first consumer card to handle Transform and Lighting calculations in hardware on the chip itself. It shifted a major workload off the CPU and changed how games were built going forward.
The NVIDIA GeForce 3 (released in 2001) introduced programmable shaders to consumer graphics. Which major gaming title was closely associated with its launch and helped showcase its capabilities?
Correct! Halo: Combat Evolved was originally shown running on the GeForce 3 hardware at Macworld 2000 when it was still a Mac and PC title, showcasing the card’s programmable vertex and pixel shaders. The GeForce 3 brought DirectX 8-class features to consumers for the first time.
Not quite. Halo: Combat Evolved was famously demonstrated on GeForce 3 hardware during its early reveal, helping showcase the card’s then-revolutionary programmable shader capabilities. The GeForce 3 was a landmark card that introduced DirectX 8 features to the consumer market.
The ATI Radeon 9700 Pro, released in 2002, is widely considered one of the most impactful GPU launches ever. What DirectX feature class did it introduce to the consumer market?
Correct! The Radeon 9700 Pro was the first consumer GPU to fully support DirectX 9 and Shader Model 2.0, and it did so while also significantly outperforming NVIDIA’s competing cards at the time. It’s often cited as one of the greatest GPU launches in history due to its combination of features, performance, and value.
Not quite. The Radeon 9700 Pro was the first consumer card to bring full DirectX 9 and Shader Model 2.0 support to the market, leapfrogging NVIDIA’s lineup. It was so well-received that it’s still talked about as one of ATI’s — and the industry’s — greatest GPU launches ever.
The NVIDIA GeForce GTX 1080 Ti, released in 2017, became a legendary card for enthusiasts. Approximately how much GDDR5X video memory did it feature?
Correct! The GTX 1080 Ti shipped with 11 GB of GDDR5X memory on a 352-bit bus, which was remarkably generous for a consumer card at the time. Combined with its Pascal architecture, it delivered near-Titan X performance at a lower price and remained competitive for years after its launch.
Not quite. The GTX 1080 Ti packed 11 GB of GDDR5X memory, which was unusually large for a consumer-grade card in 2017. This generous VRAM buffer, paired with its powerful Pascal architecture, is a big reason it remained relevant and beloved by gamers for so many years.
3dfx Interactive, maker of the legendary Voodoo series, eventually went out of business. Which company acquired 3dfx’s assets and intellectual property in 2002?
Correct! NVIDIA purchased 3dfx’s assets, patents, and intellectual property in late 2000, with the deal finalized around 2002. This acquisition eliminated one of NVIDIA’s most formidable rivals and gave NVIDIA access to 3dfx’s engineering talent and SLI patents, which NVIDIA later revived under its own branding.
Not quite. It was NVIDIA that snapped up 3dfx’s assets and patents after the company collapsed. The acquisition was a pivotal moment in GPU history, removing a major competitor and handing NVIDIA the SLI technology it would later resurrect for its own multi-GPU platform.
AMD’s Radeon RX 480, launched in 2016, caused a stir in the budget GPU market. What was its approximate launch price that made it so disruptive?
Correct! The RX 480 launched at just $199 for the 4 GB model (with the 8 GB model at $229), delivering performance that rivaled cards costing significantly more. It brought strong 1080p gaming performance to a mainstream price point and is credited with forcing NVIDIA to be more competitive in the mid-range segment.
Not quite. AMD priced the RX 480 at $199 for the 4 GB version, which caused a sensation because it punched well above its weight class. The card’s aggressive pricing pressured the entire market and is still remembered as a big win for consumers looking for affordable 1080p gaming performance.
NVIDIA’s RTX 20-series, launched in 2018, introduced real-time ray tracing to consumer GPUs. What is the name of the dedicated processing unit on these cards responsible for accelerating ray tracing calculations?
Correct! The Turing architecture introduced dedicated RT cores specifically designed to accelerate Bounding Volume Hierarchy (BVH) traversal and ray-triangle intersection calculations — the most computationally expensive parts of ray tracing. Without these dedicated units, real-time ray tracing at playable frame rates would not have been practical.
Not quite. NVIDIA built dedicated RT cores into its Turing-based RTX cards to handle the heavy math behind real-time ray tracing. While the cards also feature Tensor cores for AI-based tasks like DLSS, it’s the RT cores that are purpose-built for accelerating ray and triangle intersection calculations.
Your Score
/ 8
Thanks for playing!
This boosts one of AMD’s strongest selling points
AMD never really struggled with value
AMD’s Radeon graphics cards have long been the value pick when compared to Nvidia. In this generation, AMD didn’t even attempt to vie for the top dog position against Nvidia. Instead, it targeted the mainstream market with the RX 9060 XT and the RX 9070 XT.
Pricing has fluctuated on the flagship card, but in the last few months, that’s been due to the ongoing RAM-pocalypse. Although more expensive than the MSRP implies, the RX 9070 XT offers better value than Nvidia in pure raster performance.
Nvidia often wins on ray tracing, AI features, and DLSS in general, but AMD has always had a habit of making raw performance and VRAM look attainable. Adding FSR 4.1 to these older cards, which still pack a lot of VRAM and plenty of performance (looking at you, RX 7900 XTX), plays right into AMD’s strongest selling points.
- Brand
-
GIGABYTE
- Cooling Method
-
Active
AMD’s RDNA 4 graphics cards get the full benefit of FSR 4.1, and the RX 9070 XT is the flagship card. It can rival the RTX 5070 Ti at a much lower price point.
How does this compare to Nvidia’s updates?
Let’s address the elephant in the room
Nvidia was the first to lock DLSS frame gen to a certain generation, but AMD followed with RDNA 4, restricting FSR 4 to RDNA 4. To that end, Nvidia improved things for those who are still running older RTX cards, but the lineup and software availability is still quite fragmented (and confusing for some).
DLSS frame generation is still locked to RTX 40-series and newer, and MFG is an RTX 50-series exclusive. Meanwhile, performance on older cards hasn’t always been stellar, although most users still see the update as a net gain (which it certainly is).
That puts AMD in an interesting spot. FSR 4.1 coming to RDNA 3 and RDNA 2 makes the comparison a whole lot less one-sided than it used to be. If you own an older Radeon GPU, AMD is giving you a reason to hold on to it a little longer instead of shelling out money on upgrades. That’s pretty neat.
The ‘fake frames’ era: Why DLSS 4.5 is just a crutch for unoptimized AAA games
Upscaling and frame generation revolutionized gaming, but it’s not all good news.
This could make used and renewed Radeon GPUs a lot more tempting
This is music to my ears in these market conditions
AMD’s RDNA 3 and RDNA 2 cards still offer a lot of value, and with this new update, their prices might actually rise, so if you’re planning to buy one, now is the time.
Cards like the RX 7800 XT and the RX 7700 XT remain a solid option for AAA gaming, especially with FSR 4.1 upscaling on the horizon. Their prices on Amazon aren’t too bad, either. You can buy the RX 7800 XT for close to $550, making it a solid $150-$200 cheaper than the RX 9070 XT.
The RX 7900 XTX held up well in price, which makes it a trickier pick. It still goes for over $1,100. Although it packs more RAM than the RX 9070 XT, and actually beats it in rasterization, the extra $400 makes it a tough sell.
One way or another, digging into older AMD GPUs is once again super relevant, and that’s always nice to see.
This is what long-term GPU support should look like
This is a good move, and one I hope we see more of from both AMD and Nvidia. New GPU features don’t have to come to every old card forever, but when the hardware is still very capable, pushing upgrades via software gets tricky. FSR 4.1 may not make RDNA 3 and RDNA 2 brand new again, but it’ll make these cards a lot easier to recommend even in 2026.
Stephan is the sports journalist for the Maple Grove Report.
