Synthetic Benchmarks

Before jumping into the first set of performance benchmarks, let's touch on the RX 570 and its clock speeds for a little reference. So the reference clocks for the RX 570 are 1168Mhz, for comparison the RX 470 was 926Mhz. Now the XFX RX 570 RS that I’m checking out today though has a bit of an overclock over the 1168 clock speed with it bumped up to 1264Mhz. So with all of that in mind, I started my benchmarking with our Synthetic benchmarks to get a look at how the RX 570 RS compared to the other cards.

The first set of tests I ran were in 3DMark using the Fire Strike benchmark and the newer Time Spy benchmark. For Fire Strike, I ran all three settings that over 1080p, 1440p, and 4k and then Time Spy tests DX12 performance. I was a little surprised though to see that the RX 570 didn’t really best the overclocked XFX 470 I retested at the RX 580 launch. The gap was minimal at best but I was really expecting it to be a touch quicker given the small clock speed improvement, I can really only assume that the press driver I did the original testing on did a little better here than the current driver. It was better in Time Spy at least.

In the Unigine based benchmarks, the results were similar, with almost no different between the overclocked RX 570 and overclocked RX 570. The 570 did outperform the 1050 Ti by a big margin though.

In Catzilla I tested at 1080p, 1440p, and 4k with the RX 470 RS coming in right in the middle of all of the RX 470’s once again. I wish I had more to say here. I expected it to be slightly faster but given the 570 is just a refresh of the 470 it's not a big surprise, especially when comparing it to other overclocked cards.