Important preliminary note
We’ll do the same again today with the benchmarks as we did in the previous tests. This is important because in the sum of all the games, the peculiarities of the respective architectures quickly blur and these measurements also remain comparable to the previous articles, because I removed a few of the really big cards and added a few smaller cards in their place. There are only 10 specially selected games in the end, but I chose these as examples from over 20 titles and the pre-tests with several cards, because the result was almost exactly the same in the end. The weighting between the titles with pure raytracing without DXR and with DXR was done in a ratio of 6:4, with the four DXR titles coming out very differently.
Full ray tracing fun in Cyberpunk 2077, combined with more medium-weight effects like in Metro Exodus EE and the hybrid implementation of lighting to Marvel’s Guardians of the Galaxy and Shadow of the Tomb Raider, where ray tracing really only comes into play humanely. DXR is being implemented in more and more games, and the current engines almost all allow for it by now. From this point of view, it would be just as unfair to completely forego it as it would be to exclusively use such titles with DXR. Since every user has different preferences and some prefer to do without DXR completely (why actually?), I’ll accommodate all target groups a bit for once.
Sum of all games
First I’ll show you the normalized results of all games, because that’s what most people are most interested in. Of course, the details come right after that in the usual detail. I normalized the FPS and percentiles and formed a geometric mean (Geomean), because statistically this is simply more accurate and is also handled that way in the industry. We compare the Founders Edition with normal clock and 160 Watt power target with the card overclocked by 210 MHz and 175 Watt power target. This fully covers all “MSRP cards” and the OC cards. I wrote it at the beginning: all cards are the same and one can do both. Because it is allowed to. But then I am allowed to do that, too.
The 160 watt card is only about 1.5 percentage points behind the old GeForce RTX 3070 in Full HD. Even if the latter would have been about 2.7 percentage points faster in pure frame rates, it pays for the disadvantage in ray tracing with a much smaller gap. Overclocking the RTX 4060 Ti FE even results in a lead of almost one percent for the new GeForce RTX 4060 Ti. This is because almost 3 GHz are then also available, even under hard loads. Thus, the RTX 4060 Ti hits the performance point of the GeForce RTX 3070 in Full HD and outclasses the GeForce RTX 3060 Ti by at least 9 percentage points.
In P1, i.e. the Min FPS, the two variants with the Ada chips are very clearly ahead of the old GeForce RTX 3060 Ti and also still ahead of the GeForce RTX 3070, especially with OC. However, that is also due to the RT games and the few watts more, of which the Tensor and RT cores profile strongly. And even if it is only 5 watts more can flow into the budget here.
The power consumption turns out as expected, but the saved almost 60 watts compared to a GeForce RTX 3060 Ti and the whopping 75 watts compared to the RTX 3070 are already a real claim.
Single games and all metrics as gallery
As usual, I now also have an overview of all tested games with the most important metrics for you:
In addition, there is now the power consumption of the GPU (and CPU in combination) as well as the efficiency consideration