Nvidia GeForce RTX 2060 Assessment & Mega Benchmark

After an inevitable delay, it's finally time for our handy GeForce RTX 2060 cover. The first reviews based on the Nvidia Founders Edition card were published on January 7, the same day that Nvidia officially launched the card. Unfortunately we were not considered. Then cards appeared on the shelves on the 15th and sold for around $ 350 MSRP. By then we had received our reviews, but since we were too late technically, we decided to take the time and do a very thorough benchmark comparison instead of hurrying.

Today we have a 36 game benchmark with over a dozen different GPUs. At the end of the article, you'll also find a number of major performance issues and the usual cost per frame. When we're done, you'll know exactly where the RTX 2060 is and what it has to offer players.

For testing, we have the Gigabyte RTX 2060 Gaming OC Pro and the MSI RTX 2060 Gaming Z. Both are great looking boards, but since the Gigabyte card arrived first, we used it for all tests. Later we will look at more detailed information about thermals and other elements that are reserved for individual graphics card tests as they vary from model to model.

The RTX 2060 GPU is an interesting offer. While the RTX 2070 is available for $ 500, the RTX 2060 at $ 350 is 30% cheaper and offers only 17% fewer CUDA cores. It uses the same 14 Gbps GDDR6 memory, but the memory bus has been reduced by 25%, meaning that the memory bandwidth has been reduced by that amount. The RT cores were also reduced from 36 to 30.

Still, the RTX 2060 shouldn't be much slower than the RTX 2070 in terms of value, it's probably a better product. It is also cheap to compete with AMD's Vega 56. So it will be interesting to see how the two compare.

To continue with the benchmark results, we will discuss the most notable of the 36 games tested and then jump into all of our data in our breakdown analysis. Our GPU test system consisted of a Core i9-9900K with 5 GHz and 32 GB DDR4-3200 memory. We used AMD's Adrenalin 2019 Edition 19.1.1 drivers for the Radeon GPUs and Game Ready 417.35 WHQL for the GeForce GPUs.

Benchmarks

We start with Shadow of the Tomb Raider. Here the RTX 2060 with the highest default for game quality was good for an average of 63 fps. It was on a par with the GTX 1080 and the Vega 64 and thus only 5% slower than the RTX 2070. It was also 9% faster than the GTX 1070 Ti, but only 3% faster than the Vega 56.

In the Strange Brigade, the RTX 2060 averaged 86 fps, or 15% slower this time than the RTX 2070. It was also 7% slower than Vega 56, though that's not surprising considering how well this title runs on AMD hardware . Nevertheless, the RTX 2060 displaced the GTX 1070 Ti, so the overall performance was very respectable.

The RTX 2060 performs very well in Battlefield V and is only 4% behind the RTX 2070. It was able to beat the GTX 1070 Ti and Vega 56, so far the RTX 2060 has made the RTX 2070 pretty pointless.

The RTX 2070 retires when testing with Sniper Elite 4, here the RTX 2060 was 15% slower, but that's not a big margin as it costs 30% less. We also see that the 2060 can sit on the GTX 1070 Ti and the Vega 56.

Monster Hunter World plays pretty well with the RTX 2060 at 1440p. With the highest preset, we saw an average of 54 fps, which meant that the 2060 was barely slower than the Vega 64 and GTX 1080. It was also 4 to 5 fps faster on average than the GTX 1070 Ti and Vega 56. It was also only 11% slower than the RTX 2070.

The new Turing GPUs don't stack as well in Warframe compared to their Pascal predecessors. For example, the RTX 2070 is 8% slower than the GTX 1080, while the RTX 2060 is 6% slower than the GTX 1070 Ti and only 2% slower than the Vega 56.

It continues with Just Cause 4. Here the RTX 2060 was only a few frames slower than the GTX 1080 and Vega 64 with an average of 59 fps. It was also 11% slower than the RTX 2070 and Vega 64 Liquid. The frame time performance was similar from the GTX 1070 to the GTX 1080.

Leave a Reply

Your email address will not be published. Required fields are marked *