My Ryzen 5 1600 and RX 580 perform differently than benchmark results show.
My Ryzen 5 1600 and RX 580 perform differently than benchmark results show.
I recently upgraded my system with a Ryzen 5 1600 and Sapphire RX 580, based on a friend’s recommendation and benchmark reviews. The performance in Overwatch and GTA V was underwhelming, with average frame rates of 30-40 fps on low settings for GTA V and 40-50 fps in Overwatch, both at 1080p. These results were significantly lower than benchmarks using comparable hardware, which consistently achieved above 60 fps. I have reinstalled the drivers twice. My system specifications are as follows: CPU: AMD Ryzen 5 1600AF; GPU: AMD RX 580; SSD: Adata SU635 240GB; HDD: WD WD5000AVDS-63U7B1 500GB; RAM: Viper 3200 C16 Series 1x8GB; Motherboard: Gigabyte GA-A320M-S2H-CF.
It’s common for RX580s to not reach advertised frequencies due to hitting power limits first. Increasing the “power limit %” slider by 20-25% can resolve this. If you are obtaining frequency data from an OSD or logger, that information is accurate. Simply viewing the "target" frequency in Afterburner or WattMan (WattMan is recommended due to its comprehensive features) does not represent the actual frequency. Do you understand this point?
You possess single-channel RAM.
Is the RAM effectively operating at 3200MHz (it will not do so by default)?
How comparable are these benchmark configurations to your system?
Can you verify that your RX580 is consistently performing at 1300+MHz under prolonged operation?
The RX580 is running at 1350 MHz, but the RAM was not close to 3200 MHz. When I decided to enable XMP, which changed the memory speed to 3200 MHz, my PC would turn on for a minute or two and then crash. I don't know how to change the memory speed manually. Regarding benchmarks, the only similarities between my systems and others are the CPU, GPU, and memory size.
Has your motherboard BIOS been updated with the newest AGESA code? It may limit you to 3000MHz, depending on factors. Where is the GPU speed of 1350MHz under load sourced from?
It’s frequently common for RX580s to fail to reach their advertised frequency due to hitting power limits first. A quick solution is to raise the “power limit %” slider by 20-25%. Then, they will attain their advertised frequency. If you are obtaining the frequency from an OSD or logger, that measurement is accurate. However, simply viewing the “target” frequency in Afterburner or WattMan (WattMan is recommended due to its installation and expanded functionality compared to Afterburner) isn't the actual frequency. Do you understand what I’m explaining?
If your BIOS update enabled the CPU to run RAM at 3200MHz, you can ignore my previous statement regarding that. If you still cannot stably run RAM at 3200MHz, it is likely due to the Ryzen memory controller and you may need to run RAM at a slower speed to achieve stability. The intermittent issues you are experiencing are likely caused by using only one RAM stick instead of two for dual channel operation.