The topic Nvidia thinks my RTX 20-series GPU is too old for frame generation, but I use it… is currently the subject of lively discussion — readers and analysts are keeping a close eye on developments.
This is taking place in a dynamic environment: companies’ decisions and competitors’ reactions can quickly change the picture.
When I used Lossless Scaling for the first time in my life, I was amazed. The first thought that occurred to me was that frame generation, as a technologies, had been successfully democratized. A $7 utility on Steam had done what the incumbent semiconductor giants wouldn’t, which was, to bring frame generation to hardware they had already written off. Eight years later, I’m convinced I was right.
There’s an economic barrier that this utility shattered when it launched frame generation support, and to this day, it continues to do so. Nvidia had locked DLSS 3 frame generation behind the Ada Lovelace architecture, and AMD’s proprietary tech had its own hardware barriers. Sure enough, AMD later opened things up with Fluid Motion Frames around the same time, but the fact that a solo developer on Steam had already proven that it was possible on any GPU made a strong case for it anyway. Here’s why LS occupies a permanent place in my Steam Library.
The RTX 2070 Super is a particularly uninteresting card, at least by today’s standards. It doesn’t feature native frame generation, comes with only 8GB of GDDR6 memory, and is downright average when it comes to raw rasterization, and that’s when you compare it to cards of its own generation.
The irony is that all of these factors make it the perfect GPU for Lossless Scaling, which means that the amount of uplift in the gameplay experience is going to be most pronounced at 1080p and 1440p resolutions. Because the card can still push most modern titles to a stable 40–50 FPS at medium-high preset, it gives LSFG’s 2x mode the ideal base frame rate for the jump. The frame generator can easily work with that output and double it to 80-100 FPS, at which point, the performance is ideal for an immersive experience on a high-refresh rate monitor for non-competitive, single-player titles that can afford the latency hit. The card is not too weak to produce clean base frames, and it also isn’t powerful enough to not need a crutch. That’s the exact window where Lossless Scaling delivers the most of its value.
Perhaps one of the most disappointing aspects of DLSS is the fact that it relies exclusively on developer support for integration in titles on a per-game basis. This means if a studio hasn’t built support for the feature, you can’t leverage it. This also means that older titles, indie games, and emulation, are automatically disqualified from running frame gen.
The reason this is a particularly frustrating problem with emulators is the fact that you can almost always rely on them to be optimized poorly. The implementation of OpenGL and Vulkan on most emulators is frequently suboptimal, which is understandable as various technical and developmental challenges exist, especially when you consider that they receive very little funding. While many emulators support these APIs, it’s not at all unusual to see inconsistent and choppy performance, driver incompatibilities, and a constant need for manual tweaking.
In this regard, I’ve found Lossless Scaling to be nothing short of an absolute godsend. Because it acts as an overlay layer that functions on top of whatever is being emulated, it alleviates performance woes the moment it’s brought into action. Although it doesn’t help much with occasional introduction of artifacts, it’s the one utility that has single-handedly saved me hours worth of tweaking just to enjoy the titles I’ve ported to my setup and make them feel playable.
It should surprise no one who has been watching the GPU market that the pricing hasn’t made sense for a while. Mid-range cards are now verging into the luxury-goods territory, and the ongoing DRAM crisis is poised to push things even further in the coming years. Upgrading from a 2070 Super to anything in the current Blackwell or Ada Lovelace lineup means spending upwards of $500–$700 for what is functionally a better version of a similar experience that’s already possible with a $7 utility, at least in ways that matter to me.
You’ve probably heard me mention a dozen times in this piece that Lossless Scaling costs $7, but part of it is sheer disbelief that a technologies capable of extending the useful life of a 6-year-old GPU by years is available for the price of a cappuccino. Beyond doubt, it offers an undeniably compelling value proposition in PC gaming, and I can’t really believe it hasn’t been priced higher. But I’m not complaining.
When Nvidia decided that frame generation wasn’t coming to the RTX 20-series, it seemed like the end of the line for the card, until a single utility from a solo developer proved otherwise. The card still renders, still performs, and does everything that I need for a satisfactory gameplay experience on the hardware that I own. Perhaps the best of all is the fact that it works on any game and any GPU, which is already a little more than DLSS has ever been able to claim.
With its powerful frame generation and upscaling provisions, Lossless Scaling offers an affordable way to increase your in-game FPS.