Test Methodology
So what we have done is taken a single 32GB stick of 6000MHz CL40 DDR5 and put it up against a dual-channel kit of two 16GB sticks running at the exact same speeds and timings. Much like we did for the Ryzen 9000 series, we wanted to see if dual-channel is still the way to go, or if you could make do with a single stick that could potentially be upgraded at a later date. In theory, this would give you an even larger capacity at some point down the line when prices stabilise.
What we have done is looked at the Core i5 14600K, i7 14700K and i9 14900K at 1080p, where memory constraints typically show their ugly head, to see how much, if any, the difference is between both scenarios.
To test, we put these chips and memory into the ASUS ROG Z790 Maximus Extreme with an RTX 5090 Founders Edition to rid the results of any GPU bottleneck. For the memory, we used a single stick of Apacer NOX 32GB 6000MHz CL40 memory for our single-channel tests, and an XPG Lancer Blade 32GB dual-channel kit for our other tests. As the timings are different between the kits, we normalised the timings and sub-timings to be the same for both scenarios to give a fair test.
I will also say that while low settings will show a higher variance, we tested on higher settings, as these are where the majority of the user base will be playing anyway, so it is a truer representation of what the average gamer will see.
So with that out of the way, let us get onto those glorious benchmarks.
