This site may earn chapter commissions from the links on this page. Terms of utilise.

On Monday, Nvidia appear a new set of GPUs in a presentation focused on ray tracing and the appearance of ray tracing in today'southward games. Nvidia has congenital an entirely new capability effectually ray tracing and announced an all-new SM (Streaming Multiprocessor) architecture effectually information technology. Simply Nvidia also debuted GPUs at substantially college prices than its previous generation, and it showed no straight benchmark data that didn't revolve effectually ray tracing.

Buying CPUs and GPUsSEEAMAZON_ET_135 See Amazon ET commerce for a first-generation feature is almost always a bad idea. If yous bought an Nvidia Maxwell or Pascal video menubecause you idea DX12 and Vulkan were the time to come, practise you feel similar you lot got what you paid for as far equally that characteristic is concerned? Probably not. AMD doesn't get to take a bow on this either. True, DX12 has been kinder to Team Red than Team Greenish, simply if yous bought a 2022 Radeon thinking Mantle was going to take over the gaming industry, you didn't get a lot of shipping titles earlier information technology was retired in favor of other APIs. If you lot bought a Radeon in 2022 thinking you lot were getting in on the dawn of a new age of gaming, well, you were wrong.

The list goes on. The outset DX10 cards weren't peculiarly fast, including models like Nvidia'south GTX 8800 Ultra. The outset AMD GPUs to support tessellation in DX11 weren't all that skilful at it. If you bought a VR headset and a acme-end Pascal, Maxwell, or AMD GPU to drive information technology, guess what? By the fourth dimension VR is well-established, if information technology e'er is, you lot'll be playing it on very unlike and vastly improved hardware. The first strike against buying into RTX specifically is that by the time ray tracing is well-established, practically useful, and driving modern games, the RTX 2080 volition exist a garbage GPU. That's not an indictment of Nvidia, information technology's a consequence of the substantial lead time betwixt when a new GPU characteristic is released and when plenty games take advantage of that feature to make it a serious perk.

But in that location'due south too some reason to ask but how much performance these GPUs are going to deliver, period, and Nvidia left substantial questions on the table on that point. The company showed no benchmarks that didn't involve ray tracing. To endeavour and predict what we might see from this new generation, let's take a look at what past cards delivered. We're helped in this by [H]ardOCP, which recently published a massive generational comparison of the GTX 780 versus the GTX 980 and 1080. They tested a suite of xiv games from Crysis 3 to Far Weep 5. Permit'due south compare the GPUs to the rate of performance improvement and see what we tin can tease out:

RTX2080-Comparison

Click to enlarge

There'southward a lot going on in this chart, and then let's suspension it downwards. When Nvidia moved from Kepler to Maxwell, we see evidence that they made the core far less dependent on raw memory bandwidth (the GTX 980 has markedly less than the 780), but that this lost Nvidia nothing in overall functioning. Maxwell was a meliorate-balanced architecture than Kepler, and Nvidia successfully delivered huge performance improvements without a node shift. Merely while Maxwell used less bandwidth than Kepler did, it even so benefited from a huge increase in fill rate, and the overall comeback across 14 games tracks that fill-rate heave. Clock speeds too increased essentially. The percentage comparison information from [H]ardOCP reflects the 14 game improvement for the GTX 980 compared with the GTX 780, and then from the GTX 1080 compared with the GTX 980.

Maxwell to Kepler duplicates this improvement. Fill up rate increases a monstrous 1.6x, thanks to the increased clocks (ROPs were identical). Bandwidth surged on the adoption of GDDR5X, and the overall improvement to gaming performance is directly in line with these gains. The point here is this: While any given game may gain more or less depending on the specifics of the engine and the peculiarities of its design, the average trend shows a strong relationship between throwing more bandwidth and fill rate at games and the performance of those titles.

At present we come up to RTX 2080SEEAMAZON_ET_135 See Amazon ET commerce. Its make full rate is actually slightly less than the GTX 1080SEEAMAZON_ET_135 See Amazon ET commerce. Its core count increase is smaller than either of the previous two generations. Its bandwidth increment is smaller. And those facts alone propose that unless Nvidia managed to deliver the mother of all IPC improvements via rearchitecting its GPU core, the RTX 2080 family is unlikely to evangelize a huge comeback in current games. This tentative decision is further strengthened by the company's refusal to show any game information that didn't focus on ray tracing this week.

The Ray Tracing Hereafter Isn't Hither Yet

Furthermore, once you look at what Nvidia is using RTX for, it's articulate that the visitor isn't actually going to deliver completely ray-traced games. Instead, the focus here is on using ray tracing to handle certain specific tasks, like improved dissonance reduction or shadow work. And that's fine, as far equally information technology goes. Screenshots from PCGamesN (provided by Nvidia's comparison tools) show that RTX can make a overnice difference in certain scenes:

Right: RTX on. Left, RTX off.

Just the RTX hardware in the Nvidia GPU, including the RTX 2080 Ti, isn't going to be fast enough to merely ray trace an entire AAA game. Even if it was, current game engines themselves are not designed for this. This point just cannot be emphasized enough. There are no AAA game engines that deploy ray tracing every bit their primary rendering method. It'southward going to have fourth dimension to create them. At this phase, the goal of RTX and Microsoft DTX is to let ray tracing to be deployed in certain areas of game engines where rasterization does poorly and ray tracing could offering amend visual fidelity at essentially less performance cost.

That'due south not a new realization. When I wrote almost ray tracing in 2022, one point I came beyond is that there are certain areas where ray tracing tin can really be faster than rasterization while providing a higher-quality result. Combining the ii techniques in the same engine is tricky, keeping the ray tracing fast enough to work in real time is tricky, and Nvidia and Microsoft both deserve credit for pulling it off — just keep in mind precisely what you're buying into, here. Despite the implications of the hype train, you aren't going to exist playing a game that looks like a ray-traced version of Star Wars any time presently, because the GPU that would deliver that kind of fidelity and resolution doesn't be. Demos are always going to look better than aircraft products because demos aren't concerned with emulating the unabridged game world — just the pretty visuals.

Await to the RTX'south features to provide a nominal boost to image quality. Only don't wait the moon. And never, e'er, buy a GPU for a feature someone has promised yous will appear at a later on date. Buy a GPU for the features it offers today, in shipping titles, that yous can definitely take advantage of.

What Tin We Say Nearly RTX Performance?

I'm unwilling to declare the RTX 2080'southward operation a settled question because numbers don't always tell the whole story. When Nvidia overhauled its GPUs from Fermi to Kepler, it moved to a dramatically dissimilar architecture. The power to predict performance equally a result of comparison core counts and bandwidth broke equally a result. I haven't seen any data that Turing is as large a departure from Pascal as Kepler was from Fermi, simply it'south always best to err on the side of caution until formal benchmark data is available. If Nvidia fundamentally reworked its GPU cores, it's possible that the gains could be much larger than unproblematic math suggests.

Notwithstanding, simple math suggests the gains hither are not particularly strong. When you combine that with the real-just-less-than-awe-inspiring gains from the incremental addition of ray tracing into shipping engines and the significant cost increases Nvidia has tacked on, there's good reason to keep your wallet in your pocket and wait and see how this plays out. Just the only fashion the RTX 2080 is going to deliver substantial performance improvements above Pascal, over and above the 1.2x – 1.3x suggested past core counts and bandwidth gains, is if Nvidia has pulled off a huge efficiency proceeds in terms of how much work can be done per SM.

Now Read: How Nvidia RTX Works, Nvidia Unveils Turing GPU Architecture, and Nvidia Launches New RTX GPU Family