Amd Fsr 20 Is Live In Deathloop And Could Be A True Nvidia Dlss Rival

FidelityFX Super Resolution 2.0 is AMD’s second effort at offer a broad game upscaling engineering science to compete with Nvidia’southward DLSS. In this article nosotros’ll be using the game Deathloop which supports DLSS, FSR one.0 and FSR 2.0 to criterion and compare several GPUs from electric current and previous generations, including the GTX 1070 Ti, GTX 1650 Super, RTX 2060, RTX 3080, and on the Radeon camp, the RX 570, Vega 64, RX 5700 XT, and RX 6800 XT.

Since the release of DLSS two.0 over ii years agone, it’s steadily get a key selling point for Nvidia’s RTX GPUs, lots of gamers take been buying GeForce GPUs specifically for this feature, and AMD has been working on something to counter that for their Radeon customers.

AMD’s first attempt, at present called FSR one.0 was released about a year ago and while we found that FSR 1.0 was effective in some situations, particularly at 4K using the Ultra Quality mode, it ultimately wasn’t as adept equally DLSS overall. The image quality at 1080p was unimpressive and the Performance way looked worse than DLSS.

FSR 2.0 is a meaning upgrade to the upscaling technology, moving from a spatial technique to temporal. If you want to larn more most how FSR 2.0 works, we covered a lot of information when it was first appear and then information technology’s worth revisiting that.

The shift to temporal scaling allows FSR two.0 to combine current frame data with data from previous frames and motion vectors to reconstruct the final epitome, utilizing a lot more than data than is possible with spatial scaling. By having access to all this data, it’s possible for FSR 2.0 to produce “ameliorate than native” results, like Nvidia likewise claims for DLSS.

FSR 2.0 vs DLSS two.0: How they work

Where FSR ii.0 and DLSS differ is in how they approach the puzzle of reconstruction and choosing which bits of data to use. DLSS takes a closed-source AI based approach, which requires specific instructions exclusive to Nvidia’s Tensor core hardware.

FSR two.0 doesn’t use AI at all, instead using a paw-tuned algorithm that’s open up-source and supports a wide range of hardware, just like FSR i.0 – tensor cores or other specialized AI processing blocks are not required. Nosotros were able to get FSR 2.0 working on 5-yr-old hardware without any effect, more on that later.

The kickoff opportunity to evaluate FSR ii.0 came with Deathloop, a game which also supports FSR one.0 and DLSS, and so we could exercise a dainty comparison between today’s main upscaling technologies from the two GPU vendors. Cheque out the video below for a lot of particular on how visual quality differs betwixt FSR 2.0 and DLSS in Deathloop.

At present I wouldn’t telephone call this an FSR 2.0 “review” only because we’re looking at a sample of one game. We’ll need more than that to brand a definitive call on FSR two.0 vs DLSS, but at the very least it’s a good preview of what FSR 2.0 can do.

Deathloop is also a skillful game to exam with because DLSS is very effective in this title. It’south 1 of the all-time examples of DLSS producing a “ameliorate than native” image, while FSR 1.0 looks less than amazing.

FSR ii.0 Settings

FSR ii.0, similar DLSS, has three quality modes available in Deathloop: Quality, Balanced and Performance, which stand for to a one.5x, 1.7x and 2.0x scale cistron respectively, roughly in line with what Nvidia offers. This means that at 4K, using FSR ii.0 Quality means we’re upscaling from 1440p to 4K, while using the Performance mode gives usa a 1080p to 4K upscale.

FSR ii.0 is a widely supported temporal upscaling solution, pregnant it can work across a range of hardware and that’s exactly what we’ll be doing today. Nosotros’ve tested GPUs from four generations of AMD and Nvidia releases to run into how well FSR 2.0 scales on different architectures.

For today’s testing, we’re using a Ryzen 9 5950X test arrangement equipped with 16GB of low latency DDR4-3200 retentiveness, running the latest publicly available drivers for AMD and Nvidia. Nosotros’ll be comparing FSR ii.0 to the side by side best available upscaling solution for the GPU at hand, and so for RTX 20 series cards and newer we’ll be comparing to DLSS, and everything else to FSR one.0.

To foreclose CPU bottlenecks, nosotros’ve run Deathloop at the best quality playable settings for each GPU, which includes ray tracing enabled on higher stop cards. Deathloop is a enervating game on GPU memory, and some of the cards nosotros’ve tested have only 4 GB or 6 GB of VRAM, which causes issues on the highest settings and basically bottlenecks the carte – significant we don’t meet the real benefits of FSR 2.0 or other upscaling algorithms. For those GPUs, we’ve reduced the settings to an appropriate level that doesn’t crusade a bottleneck.

Popular:   Microsoft Defender For Business Now Available As Standalone Product


Radeon RX 570: FSR one vs. FSR 2 vs. Native

We’ll start hither with the oldest and slowest card in the line-upward, AMD’s trusty Radeon RX 570 4GB, which was released style dorsum in 2017. This game struggles to run Deathloop at 1440p using even Medium settings, but with upscaling we can get a operation uplift at a minimal toll to visuals.

The RX 570 doesn’t benefit from FSR 2.0 to nearly as significant of a degree. We practise get a performance uplift, only it’s simply a ten percent gain from the FSR ii.0 Quality mode, and 27 percent from the Performance mode.

It’due south yet worth using, just non the instant 40%+ gains we saw from the latest architectures. You tin can also see that FSR one.0 is indeed faster, which wasn’t the case with RDNA2 either. With the 6700 XT, typically FSR ii.0 Quality way ran amend than FSR ane.0 Ultra Quality. Merely with the RX 570, it’s the less taxing FSR 1.0 that runs a few frames better. However, I’d withal recommend using FSR 2.0 here every bit the visual quality is significantly superior at 1440p.

At 1080p, FSR 2.0 was more capable of a functioning uplift. The Quality mode was giving a 14% boost over native rendering, and the Performance mode (which nosotros don’t really recommend at this resolution) was 26 percentage faster. This does help the RX 570 accomplish an even more playable frame rate, but it’due south articulate the gains from this old, mainstream GPU are limited.

GeForce GTX 1650 Super: FSR ane vs. FSR 2 vs. Native

The GeForce GTX 1650 Super is besides an entry-level GPU but information technology’s a off-white bit newer but doesn’t support DLSS equally it lacks tensor cores. Like the RX 570, the GTX 1650 Super doesn’t benefit hugely at 1440p with either FSR two.0 or i.0 in this title, and it seems that even Medium settings is a scrap of a stretch here.

The FSR 2.0 Quality mode was only 6 percent faster than native, and the Operation mode barely improved upward on that. In comparison, FSR ane.0 was able to evangelize much better frame rates, with the Quality mode there delivering higher FPS than FSR ii.0 Functioning.

However the gains at 1080p were more acceptable. The FSR 2.0 Quality fashion was xv pct faster than native, like to what was seen on the RX 570, and the Performance mode increased that figure to 29 percent. Not earth shattering results – and FSR one.0 is definitely faster on this entry-level GPU – but usable. I’ll be interested to see how budget GPUs like this fare in other FSR 2.0 games to see if information technology’s more of a game thing or more than of an FSR 2.0 algorithm matter, but certainly it seems in that location only aren’t that many GPU resources to go around to run something similar FSR two.0.

Radeon RX Vega 64: FSR i vs. FSR 2 vs. Native

Permit’southward become back in fourth dimension to wait at FSR ii.0 running on the Radeon RX Vega 64, some other 5 yr old production. Using Ultra settings at 1440p, FSR 2.0 did provide a modest performance uplift, 25 percent for the Quality manner compared to native rendering, which was enough to take the game upwardly to a 60 FPS average.

Unlike with the entry-level cards nosotros were just looking at, we’re back into a situation where the FSR two.0 Quality mode performs better than FSR 1.0 Ultra Quality, and for that reason FSR two.0 is definitely the preferred option due to its much better epitome quality.

At 1080p, I actually saw less of a gain than at 1440p for FSR 2.0 every bit it seems like pure shading isn’t necessarily the main limiting gene for performance on Vega 64. If shading performance isn’t the principal bottleneck, then lowering the return resolution may only have express gains, which seems to be happening here. Other areas to performance, like retention or geometry, could exist holding us dorsum from further gains. But either way we still do get a performance uplift.

GeForce GTX 1070 Ti: FSR i vs. FSR 2 vs. Native

On Nvidia’due south Pascal compages we see a like situation to Vega 64. Using Ultra settings at 1440p, the GTX 1070 Ti was able to attain 20 percentage better performance going from native rendering to FSR 2.0 Quality, and 41 percent using FSR 2.0 Performance. FSR one.0 performs better on this architecture, especially at lower render resolutions, but I’d all the same prefer to apply FSR 2.0 due to its higher prototype quality.

At 1080p, one time once again more modest gains of fifteen percent for the Quality fashion, similar to Vega. Information technology’ll be interesting to run across how this holds upwards in other titles, merely FSR 1.0 wasn’t exactly miles better for something similar its Ultra Quality setting. So I’m still pleased that FSR 2.0 is usable here and a better option than AMD’due south older FSR version.

Popular:   Amd Graphics Cards Could Get A Free Speed Boost

Radeon RX 5500 XT 8GB: FSR 1 vs. FSR 2 vs. Native

Let’s take a look at a beginning-generation RDNA product, the RX 5500 XT 8GB. At 1440p, FSR 2.0 Quality mode was able to deliver 24 percent better performance than native rendering, a much larger uplift than with the RX 570 despite both cards running at approximately the same FPS natively under the conditions we tested. 45% better performance was besides possible using the Functioning mode, and overall this is preferable to FSR 1.0.

At 1080p I too saw respectable gains, 21 percent for FSR 2.0 Quality fashion and 37 percent for FSR two.0 Operation, again delivering results that are preferable to using FSR i.0. Yes the FSR i.0 Performance setting gives you a few extra FPS, but your eyes won’t forgive you for using such a depression quality setting.

Radeon RX 5700 XT: FSR 1 vs. FSR ii vs. Native

Under the same architecture family, allow’s now look at the much faster RX 5700 XT. With the same compages and same memory capacity at hand, FSR two.0 conspicuously does benefit from more GPU resource. At 1440p using FSR 2.0 Quality mode, we saw a 34 per centum performance uplift, college than the 24% we saw for the 5500 XT. A larger uplift was also possible for the Performance mode, 61 percentage hither compared to 45% for entry-level RDNA.

At 1080p we are able to get a 27 percent functioning uplift compared to native rendering using FSR 2.0 Quality fashion, and a 44% uplift using the Functioning style. With these more than mid-range products it really doesn’t brand sense to use FSR i.0 equally the Ultra Quality style runs slower than FSR two.0 for over again, worse image quality.

GeForce RTX 2060: FSR ii vs. DLSS vs. Native

Now let’s have a look at Nvidia’due south Tensor core equipped Turing architecture starting with the RTX 2060. With this GPU, FSR 2.0 functioning is similar to what we see with the RTX 3060 Ti (beneath), in that DLSS is slightly faster than FSR 2.0 overall, but not significantly so.

FSR 2.0 Quality manner was capable of a fairly unimpressive 18 percent functioning uplift, but DLSS Quality fashion only improved that to a 21 percent gain, and then bit much of a muchness. The best results for DLSS were the Performance mode, which ended up 6% faster than the equivalent FSR 2.0 setting.

And so at 1080p, like story. Only a 13 percentage functioning uplift for FSR ii.0 Quality mode compared to native, with DLSS Quality mode delivering a nineteen per centum uplift. At 1080p this does make DLSS the favorable selection every bit it also tends to wait amend, but neither option is delivering a huge functioning increment – though I’d certainly take information technology if I was struggling to hit a playable frame rate.

GeForce RTX 2080: FSR 2 vs. DLSS vs. Native

Some of the most interesting results are these with the RTX 2080. At 4K, FSR 2.0 was capable of a 26 percent performance uplift using the Quality mode compared to native, still the gains with DLSS were much larger at forty percentage for the equivalent mode.

On this GPU, DLSS was ten to 12 per centum faster at 4K, which is a significant margin and conspicuously makes DLSS the favorable selection. Over again we don’t know how FSR 2.0 works exactly, then we don’t know which sort of GPU resources information technology favors, but I think the results on the 2080 are a petty disappointing, specially when DLSS is very effective.

It’due south a like situation at 1440p. The functioning uplift from FSR 2.0 Quality was 18 percent compared to native rendering, merely DLSS Quality was achieving a 26 percentage uplift. Not as large of a difference in favor of DLSS, but Nvidia’due south technique was still around seven per centum faster overall, then once again for 2080 owners information technology would make sense to use DLSS.

GeForce RTX 3060 Ti: FSR 2 vs. DLSS vs. Native

Radeon RX 6800 XT: FSR one vs. FSR 2 vs. Native

Permit’due south accept a look at some modern high-end GPUs. At 4K with the RX 6800 XT, FSR 2.0 Quality mode was effective, delivering a 28 per centum functioning gain over native rendering. Using the Performance mode increased that to 61 percent over native, and with almost modes either matching or chirapsia FSR 1.0 it’s clear that on a high-end card like this playing a high resolution you lot should be using FSR two.0.

Radeon RX 6700 XT: FSR 1 vs. FSR ii vs. Native

GeForce RTX 3080: FSR 2 vs. DLSS vs. Native

So for the RTX 3080 using the aforementioned settings at 4K, nosotros run across a larger gain of 38 percentage for FSR ii.0 Quality vs native, and 74 percentage for the Functioning mode. However similar with other Nvidia GPUs that we’ve benchmarked, DLSS is the faster pick, delivering iv to 6 percent higher frame rates at the equivalent quality settings to FSR 2.0. So my full general thoughts on using DLSS over FSR 2.0 on Nvidia’south latest GPUs holds true.

Popular:   Iphone 14 Rumor Points To A Price Bump But For A Good Reason

What We Learned

It’s clear that FSR ii.0 runs best on the latest GPU architectures like AMD’s RDNA and RDNA2 designs, as well as Nvidia’s Ampere. These GPUs consistently delivered the best results in terms of a performance uplift compared to native rendering. We don’t recall it’s unreasonable to look minimum gains of thirty to xl percent using FSR two.0 Quality mode on modern GPUs.

In all situations it still made sense to use or at least consider using FSR 2.0 as we didn’t meet operation go backwards, and results were more often than not favorable compared to FSR 1.0, just there are some instances where gains are smaller than expected.

Nevertheless it’s non just the GPU architecture that influences performance. Like we saw several years ago when testing DLSS, a key component is also the native rendering frame rate: y’all’ll get higher gains if your baseline FPS is lower.

This is because FSR two.0 has a stock-still rendering toll, which takes up a larger proportion of the total frame rendering fourth dimension at college frame rates. Simply even if your baseline is 100 FPS or more, FSR 2.0 can still provide solid gains on newer architectures, provided you don’t run into CPU bottlenecks.

It also seems to be the instance that FSR 2.0 runs better on GPUs that are simply faster overall, every bit we saw when comparing the RX 5700 XT and RX 5500 XT: same compages, aforementioned settings, all the same the gains from the 5700 XT were better despite a higher baseline FPS. This is also what AMD suggested at launch: a longer FSR 2.0 processing fourth dimension for less powerful GPUs.

It makes sense that cards with more resources could run the algorithm faster, especially equally it doesn’t utilize defended hardware. When the algorithm is run faster, it uses up proportionally less of the time it takes to render each frame, which can requite a performance gain advantage.

This is also relevant for budget and older GPUs. Cards like the RX 570 and GTX 1650 Super are only taking longer to process FSR ii.0, which leads to more than limited performance uplifts even in “ideal” weather for upscaling. Compounding this are the limitations of older architectures, for example if parts of FSR ii.0 uses FP16 processing similar FSR 1.0, and then the algorithm will have to fallback to FP32 processing on GPUs architectures like Polaris that don’t natively support FP16, hurting performance. Nosotros don’t yet accept the full picture on what architectural features are necessary for maximum FSR 2.0 performance, only nosotros’d exist surprised if FP16 wasn’t a gene.

Having that said, even five year old GPUs tin can run and do good from FSR 2.0, which you can’t say for DLSS. We’ll certainly take a 10 to xx percent functioning uplift from this temporal algorithm on those cards, and would use it over FSR 1.0 despite lesser functioning gains.

Across the four GPUs we tested today that also support DLSS, from both Turing and Ampere generations, FSR 2.0 performance gets close to DLSS using equivalent quality settings. However, DLSS typically does run ameliorate, up to 12% faster in the best case scenario on the RTX 2080. Then for Nvidia GPU owners with Tensor equipped cards, our recommendation would be to use DLSS, for the small-scale operation advantage and also because in some situations information technology delivers better image quality too.

Finally, it’s worth repeating that this is a sample size of one game. We’ll need to practice more in depth testing a few months down the rails to see how FSR 2.0 works across a wider range of titles. Past then we should also accept the FSR 2.0 source code and hopefully experts in that area tin dive in and requite united states of america a good film of what’s going on. But for now, nosotros tin can still learn a bit about the optimal situations for FSR two.0 in Deathloop, then hopefully this testing has been valuable.

Shopping Shortcuts:
  • Radeon RX 6950 XT on Amazon
  • Radeon RX 6750 XT on Amazon
  • GeForce RTX 3080 on Amazon
  • GeForce RTX 3070 Ti on Amazon
  • GeForce RTX 3060 on Amazon
  • Radeon RX 6700 XT on Amazon
  • Radeon RX 6600 XT on Amazon

Amd Fsr 20 Is Live In Deathloop And Could Be A True Nvidia Dlss Rival