Performance
Performance has been tested on the AMD Ryzen platform, which contains the Ryzen 5 Pro 4650G processor, MSI X570 Unify motherboard, and MSI GTX1660Ti graphics card. Used OS is Windows 10 x64 and the latest updates.
If anyone skips previous explanations, I will add that the Ryzen 4000 series processors can run at much higher infinity fabric frequency. In our case, it’s up to 2266MHz, so DDR4-4533 with two memory modules and 2200MHz, so DDR4-4400 with four memory modules. Simply just right for the tested Patriot Viper 4 Blackout memory, we won’t see significant performance drops caused by something other than the memory settings. DDR4-4933 result has been made on the asynchronous frequency. Its performance is about as high as DDR4-4866, and memory timings are the same, so the DDR4-4866 was skipped.
Our comparison includes overclocking results and settings at which the Viper 4 Blackout DDR4-4400 was stable. Our motherboard’s maximum stable settings were DDR4-4933 CL18-17-17 at 1.57V for two modules and DDR4-4400 CL18-22-22 1.45V for four modules. Since four-module performance was about the same at XMP and manual settings, the manual settings were skipped.
As usual, we will start with AIDA64 Cache and Memory benchmark, which is probably the best software for synthetic memory speed tests.
AIDA64 Memory and Cache benchmark likes high memory frequency. It’s clearly scaling well with memory frequency up to the DDR4-4933, which was our maximum in tests. These results are sometimes misleading, and you will find out in a minute why I said so.
Suggested performance in memory bandwidth does not match our results in PCMark 10, where the DDR4-4133 was the slowest one. In PCMark 10, it’s the best setting, and there are two reasons why. The DDR4-4133 could run at tighter timings of 16-16-16-36 and tight sub-timings, which are helping a lot. It’s also a four-module setup that works like dual-rank, which additionally helps in this benchmark.
On the other hand, the XMP performs well, and both results at DDR4-4400 are about as high in PCMark 10.
In Cinebench series benchmarks, results are close to each other, and we can’t clearly say if one setting is better or it’s a margin error.
It’s time for some 3D benchmarks from UL(previously Futuremark).
3DMark and VRMark series benchmarks show similar results at all settings, as we’ve seen in PCMark 10. All our settings provide high results, but 32GB DDR-4133 CL16 and DDR4-4400 XMP seem slightly better than others.
The difference is so low that I doubt it would really matter and be a reason to pick the 32GB set up over the 16GB. On the other hand, if we need more RAM capacity, it seems like a good option to keep the same performance slightly better.
More demanding 3D tests at the display resolution up to 8K are not much different. However, in the Final Fantasy XV benchmark, we can clearly say which one is the fastest, while in the Superposition, it’s not so clear, and all memory settings seem about as fast.
Results in newer games like Shadow of the Tomb Raider or Far Cry 5 show us something else than we’ve seen in other tests. These games run the best at higher memory frequency, reaching the highest FPS at DDR4-4533 CL16 and DDR4-4933 CL18.
Our comparison doesn’t clearly say which setting is the best as each benchmark is different, and memory affects it differently. However, we can say that Patriot did good work with the XMP profile, even though at first sight, the main timings don’t look the lowest. Of course, we can try our luck in overclocking, but the results cannot be as high as we expect. I will tell some more about overclocking on the next page of this review.
