Performance
Tests were performed on the Intel Coffee Lake platform with the i5-8600k processor. All results are based on the HyperX Fury DDR4-2933 memory kit. The comparison includes overclocking results so we can already compare all available settings. I will only add that in 3D benchmarks has been used EVGA GTX1060 SC ITX graphics card.
The first test is as usual AIDA64 memory and cache benchmark.
As we can see, memory profiles are not so fast but are still good enough for most users. On the Intel Coffee Lake platform, it looks better than on the AMD Ryzen where everything is affected by memory speed.
The similar situation is in all AIDA64 tests so memory read, write and copy.
We have to remember that if we use lower Intel chipsets then the memory frequency will be limited to DDR4-2666 and HyperX Fury DDR4-2933 will offer the same speed as the most expensive memory kits.
Results in synthetic memory tests are often misleading so we are adding other tests to the comparison.
As you can see below, in rendering memory speed is not so important as the memory capacity. Benchmarks like Cinebench R15 will show not much better results using more expensive and high-frequency memory kits.
In Pi calculations, there is a significant difference between memory settings. On the other hand, results at DDR4-2933 are still high.
In 3D tests like the latest VRMark, there is barely any difference between memory settings. Even though we know that high frequency should help in games then at least 3D benchmarks, we won’t really see a significant difference. 3DMarks are using memory speed more for physics calculations than the graphics itself.
Below are results in two other VRMark tests so Cyan and Blue room.
The Blue Room just for comparison.
At the end SPECwpc v2.1 compression and decompression benchmark based on 7-Zip which is a free application.
In this test, we can clearly see that having faster RAM will save us a couple of seconds in compression and decompression of various files. Of course the larger the archives, the more time we will save. DDR4-2933 settings are still showing good results in this test. On the other hand, most users won’t see any special difference if they have to wait a couple of seconds more or less.
Overclocking
Disclaimer: Overclocking is never guaranteed, so the results may vary depending on certain conditions and various hardware configurations. I am not recommending overclocking if you do not know what are you doing. High voltages may damage hardware and it will not be covered by warranty.
Above you could see results made on the Intel Coffee Lake platform. Below is one result on the AMD Ryzen. HyperX Fury has really high overclocking potential and even though main timings aren’t really tight then benchmark results are showing some nice numbers.
DDR4-3733 on a DDR4-2933 memory kit looks impressive, especially that in use is AMD Ryzen platform which is known to make problems above DDR4-3200 settings. Worth to mention is also that for overclocking were used all four memory modules.
I will only add that used motherboard officially supports memory up to DDR4-3466 and even Samsung based memory kits have problems to run at DDR4-3600 or above.
