Today we are starting a new series focused on CPU cooler reviews. This series will cover different set-ups, from lower end requirements to higher end and we have standardised our build to keep results as true to life and consistent as possible. We will start our adventure with an air cooler from PcCooler, the CPS RZ620. This is their first entry to the high-end coolers and it's priced at 75 USD.
While this may be a first for PcCooler - I think it’s not a new cooler. More on that a bit later on.
We will break out the video into three sections: first, we’ll cover the physical aspects, then performance and wrap up with feedback. Timestamps will be provided, so feel free to jump to the section that interests you the most.
Let’s first go through what you get for your money and right off the bat I am really impressed with the packaging and contents. The cooler is well secured in the box with plenty of protection. Inside, you'll find accessories compatible with both AMD and Intel systems, thermal paste, and a handy in-line fan controller offering three speed modes: 1800 RPM, 2200 RPM, and a PWM range from 500 to 2000 RPM.
I also really like the included simple manual. It covers everything you need to know, starting with the crucial step of removing the plastic protective film from the bottom of the cooler – something I still occasionally overlook.
While I would have preferred pre-applied thermal paste for convenience, applying it yourself isn't overly complicated.
This cooler features a design with six heat pipes, each 6 mm in diameter, and comes equipped with 120mm fans. Its total height of 158 mm ensures compatibility with most cases.
The installation process is straightforward. For our setup, we used an AMD AM5 socket, which required inserting screws and then attaching the bracket. The clip bracket has a little arrow indicating which way it should be pointing. The arrow leads towards the CPU. You do need to hold the bracket as it has a tendency to lean down. Once the bracket is in place, apply the thermal paste and then securely screw on the cooler.
For the fan installation you will need to use the little metal clips which personally I am not really a fan of as I tend to hurt myself when using them, but it is a cost effective solution. Fans here actually have a few nice features. The pretty standard anti vibration pads on the corners as well as orientation guides. These are useful for new builders so they know which way the airflow will go. The fan cables can be daisy chained and I recommend doing so before installing them to the cooler and then connecting them to the motherboard. It will save you some hassle with cable management later on.
When it comes to RAM clearance - our testbench is using 43mm RAM and it tightly fits over the top without adjusting the fan, but since it's using metal clips - if needed you can always adjust the fan up. Do bear in mind that it will reduce airflow and thus performance.
This brings us to the cooler's performance, where we have a series of graphs to share. For tests we will be using three different power profiles: low, medium and high. The low profile is approximately 77W which represents either lower end chips or higher end chips with lower core utilisation like some single threaded games.
Medium is approximately 90W and it represents a more mid range chip or for example the new AMD non X chips at stock. And lastly we have the high profile, pushing about 220W, which represents high end systems handling multithreaded workloads. We decided not to go to the extreme high for the time being as it requires a completely different setup.
With this in mind let’s jump into the results. Our comparison includes a variety of coolers, ranging from the stock AMD Wraith to more high-end options like the be quiet! series. In the low-power test, with noise-normalised fans set to 40 dBa, most coolers hover in the 25-26°C range above ambient temperature, with the stock cooler being a notable exception. This load is clearly not a problem for these coolers, but let's check out the same test with fans set to 100%.
When blasting the fans, here we can see the RZ620 emerges as a standout, achieving the lowest temperature at 24.1°C above ambient, but also recording the highest noise level at 46.3 dBa. This performance aligns well with its design, which essentially brute-forces air through the cooler. However, I'd say this is above my comfort level, and I'd advise against setting the fan speed this high under normal circumstances.
If we look at the rest of the coolers, we see they perform only slightly worse but at considerably lower noise levels, especially be quiet! Dark Rock Elite with quiet mode enabled. That is almost half as loud, but is also about 40 USD more expensive.
Next up is 90W power load and here with noise normalised fans we see slight shift in positions and now RZ620 is sandwiched between two be quiet! towers, both of which manage to stay below the 40 dBa noise floor. The temperature difference between the RZ620 and the top-performing cooler is less than 1 degree, indicating a relatively minor disparity in cooling efficiency.
Turning the fan speed up to 100% brings the CPS cooler back up to the top and it is essentially tied with Dark Rock Elite in Performance mode, but it's worth noting that the differences in cooling efficiency among these coolers are quite minimal. Now, let's ramp things up to the more demanding 220W power load and see how these coolers manage under a real-world, high-intensity workload.
And surprise, surprise - we find that these coolers are all experiencing thermal throttling, which is understandable. While some of them are rated at 260W or even 280W, that does not always mean they can hold that performance. Several factors come into play, such as the heat distribution on the CPU. Given the CPU's small size, it creates a concentrated hotspot, making it challenging for the cooler to effectively dissipate the heat. Additionally, the conditions under which each manufacturer tests their coolers to arrive at these wattage ratings are not always clear.
This is precisely why we conduct our own tests – to provide real-world insights. Let’s dig deeper to see which cooler performs better as we can still do that.
First way to check this is CPU frequency and here we have zoomed in on a graph where CPU tries to boost past 5.1 Ghz and immediately drops down. The AMD Wraith Prism cooler is a clear loser here. We are leaving a bunch of performance on the table, but the rest of the coolers do a reasonable job. Have a look at the 3rd quadrant where you see the Dark Rock Elite using performance mode slightly edges out its competitors, but only by a narrow margin.
To better quantify which cooler is most effectively dissipating heat, we turn to our average power graph. This graph is particularly useful since all the workloads are identical. It shows the average power consumption during the benchmark. The logic here is straightforward: the highest performing cooler would experience the least amount of thermal throttling and, as a result, sustain a better load. In this example, in the noise normalised test RZ620 came in lowest out of all of the big tower coolers with about 6W difference to the cooler from Montech.
When we increase the fan speed to maximum, the performance gap narrows, and the RZ620 positions itself in the middle of the pack. It is still a significant improvement from the stock cooler but becomes a bit of a mixed bag when it comes to the rest of the coolers, until you check the pricing. The be quiet! coolers, although offering superior performance, come with a significantly higher price tag. On the other hand, the Montech cooler, delivering a comparable performance to the RZ620, is actually slightly more affordable, which leads us to our opinion.
I think PcCooler has made a good attempt at a mid range dual tower cooler. It competes well with more expensive alternatives in terms of cooling efficiency, although it does fall short in the acoustics department. For users who aren't constantly pushing their PCs to the limit, this cooler would be a suitable choice. Its installation process is simple and straightforward, complemented by sufficiently detailed instructions.
However, there's an elephant in the room. The striking resemblance to DeepCool’s AK620 cooler, but I will leave that to the companies to fight about.
What do you guys think about this cooler, would you consider getting it?