For testing, I did decide to update out test suite from the original 1800X,1600X, and 1300 from the original Ryzen launch to the more recent 3800X, 3600X, and 3400G. This meant tossing out all of the previous results so I brought in two other 120mm fan coolers as well for comparison with the Dark Rock Slim. One was the always popular Hyper 212 Black Edition from Cooler Master which gives us a look at a cheaper and well selling cooler and then the NH-U12S from Noctua which is the most expensive cooler tested but is similar to be quiet! with their focus on noise. Both coolers are a little thicker than the Dark Rock Slim, frankly there aren’t many tower coolers to directly compare the slim design with. Anyhow for testing I tested the cooler with all three CPUs across a few different workloads.
Two of those were using AIDA64’s Stress Test. I’ve found that their FPU test is ultra-demanding well beyond any real-world workload for a worst-case result then the CPU workload is a lot more realistic for most real CPU usage. Because of how demanding the FPU test is, I also track the CPU clock speeds. This is because Ryzen CPUs with stock settings like to reach temperature and just downclock the clock speeds. With the newest Ryzen CPUs I have found that the temperature isn’t as locked in, but depending on the cooler it will still downclock so keeping an eye on that is important. So in the table below I have the FPU results, how did the Dark Rock Slim do in the worst-case testing? Well the thinner cooler did fall behind the two other coolers a little. This included tests with the fan turned all the way up to 100% as well, but you can see the performance improvement when cranked up. As I found in my Threadripper cooler testing, motherboard fan profiles tend to not take into account the throttling points on Ryzen and they don’t reach 100% fan speed early enough.
The FPU tests show a situation that I don’t think happens where the CPU workload tests are what I would consider to be more realistic for using most programs and games. You can see just how much the temperatures drop in this test and here the Dark Rock Slim is running at least close to the two other larger coolers. You can also see that here 100% fan speed results show a much bigger gap as well and clock speeds aren’t a concern as well.
Because Blender is popular I thought I would also add it in as real-world workload. You can see that it is very demanding, coming in well above the CPU workload in AIDA64 but still below the FPU workload as well. But because it was demanding enough I did track clock speeds again and you can see that at least with the 3900X there was a small clock speed difference between coolers. The Dark Rock Slim was once again a little undersized compared to the larger coolers but beyond being a degree or two higher in temperatures it ran with the other coolers, especially on the mid-range CPUs with the 3600X and 3400G.