Corsair Dominator Platinum DDR4-3400 16GB Review
For our memory overclocking tests we are usually interested in two main elements: how well the memory scales with additional voltage and how versatile it is at overclocking with different timings. Though we suspect that many of the overclockers who buy this kit won’t hesitate to use 1.45V or even 1.50V, we are sticking with 1.40V to stay inline with all our previous DDR4 reviews. This allows us to compare kits on a roughly equal footing, and allows us to put all our efforts on testing timings scalability instead.
In order to make sure that there weren’t any possible CPU-related bottlenecks, the CPU Cache Voltage was set to 1.35V and the CPU System Agent Voltage set to 1.25V. We focused on four basic timing configurations (12-12-12 / 13-13-13 / 14-14-14 / 15-15-15 / 16-16-16 / 16-18-18) all with a 1T command rate for optimal performance.
With the pleasantries out of the way, let’s get to the fun stuff.
Before we start getting into how unbelievably good the above results are, click here to get a popup showing the overclocking results that we achieved with the G.Skill Ripjaws 4 DDR4-3200 kit, which was G.Skill’s flagship model up until very recently and the very best Hynix MFR-based DDR4 memory kit that we have reviewed until today.
If you compare them side by side, you will see that this new Corsair Dominator Platinum DDR4-3400 memory kit absolutely demolishes the G.Skill across the board. At the time, we sincerely believed that we had a pretty firm grasp on the upper limit capabilities of Hynix MFR memory chips, but clearly that was not the case. This DDR4-3400 is so much better than anything else that we have tested due a exceptionally stringent binning process. This kit is manufactured with the top 0.5% of all Hynix MFR ICs that Corsair have binned.
As you might expect or even rightfully demand at this price, at every timings level it achieved new records for a Hynix-based kit. While there isn’t a huge difference when it comes to the tighter timings (12-12-12/13-13-13) that Hynix never really excels at, when we reach 14-14-14/15-15/16-16-16 the gap ranges from huge to unbelievable.
This kit is so good, that ironically it looks like it might have outpaced our processor. If you notice the minimal difference between 16-16-16 and 16-18-18, and similar lack of improvement between 16-18-18-1Tand 16-18-18-2T, it definitely seems like our particular Core i7-5960X has hit a wall at the voltages we outlined above. Throwing more voltage at it and more importantly switching to sub-zero cooling would definitely help extend its limits considerably.
These screenshots are just to prove that we did indeed achieve the overclocks listed, and that they were stable enough to pass a series of mainstream benchmarking and stress testing applications. If you are doing super critical work, then maybe a little Prime95 stress testing should be done as well, but for gaming and day-to-day tasks our testing is more than sufficient.
At this point, we usually try to find the maximum stable frequency that we can achieve using a memory kit’s default timings, which in this case are 16-18-18-2T.
We kind of ruined the surprise above, but there wasn’t much improvement when going from 16-18-18-1T to 16-18-18-2T. The increase from DDR4-3441 to DDR4-3452 is obviously not enough to warrant the switch from a 1T to 2T command rate, so we decided to exclude it from our already cramped benchmarking charts.
Check out the next few pages to see our benchmarking results. We kept CPU frequencies as clock to 3.7Ghz as possible – to match the core clock set by the primary XMP profile, and increased the Uncore frequencies to around 4.2-4.25Ghz to try an eliminate any memory bandwidth bottlenecks. It’s not perfect, since there is only so much you can do with a finite number of frequency multipliers but what you will be seeing are the basically true performance differences caused by the different memory settings.
Sorry, we couldn't find any posts. Please try a different search.