Performance
Performance has been tested on the AMD Ryzen platform, which contains the R9 3900X, 12-core processor, ASUS Maximus VIII Impact motherboard, XFX RX5600XT THICC III Ultra graphics card, and Silicon Power P34A80 1TB SSD with installed Windows 10 x64, and the latest updates.
Our comparison includes XMP settings and any other settings at which the Silicon Power XPower AirCool DDR4-3200 was stable. It also includes all overclocked settings at which you can expect this memory to work. Of course, overclocking is not guaranteed, and we are adding these results as a comparison.
As usual, we will start with the AIDA64 Memory and Cache benchmark, which is probably the best application to check memory bandwidth and latency.
AIDA64 bandwidth results are not much different from what I was expecting. I wasn’t expecting such a high difference between DDR4-3600 and DDR4-4133 or higher. However, this is only a synthetic bandwidth test, so it’s better to focus on benchmarks that are showing us the comparison, which is closer to the “real-life” workload like PCMark10 below.
PCMark 10 results are confirming that the difference between XMP settings at DDR4-3200 and anything higher is minimal. We can see that in the Productivity test, XMP and DDR4-3600 settings are even better than anything below DDR4-3000 or results at DDR4-4133 or higher where the infinity fabric works at 1:2 divider.
Rendering benchmarks like Cinebench R20 are not affected much by memory performance. We can see some differences, but they’re not high enough to be seen during daily work.
About the same situation, we can see in 3D benchmarks from UL like 3DMark or VRMark. Even though there is a little performance gain because of a higher memory frequency, then it can be noticed. We wouldn’t see more than 1-2FPS if it was a game.
In VRMark, it’s even less. All memory settings provide the same performance. There are some exceptions which are showing up to 1FPS difference.
In benchmarks based on popular games, it’s more visible, but also here we can expect up to 3FPS difference between the slowest and the fastest result. The XMP is somewhere in the middle, but most users won’t care to overclock memory to gain 1FPS in games were are already over 100FPS.
Our results are showing that even without further overclocking, the AirCool memory should meet gamers’ expectations. Results are high in all benchmarks, so we are satisfied.
