Contact Information


Drop me an email

Deepcool AK620 Review

Deepcool has been doing decent when it comes to making super affordable performance air coolers. So when I got to know that they have recently launched a dual tower, dual fan setup air cooler, I really couldn’t wait to test it and check how it fares against the Ryzen 3900X in my primary system. 

I am someone who just loves the idea of having air coolers around which can perform as good as a 280 or 360mm AIO. So the Deepcool AK620 cooler was sent to be my Deepcool and I was quite thrilled to test it as soon as I got it. So I shoved down the other content and started focusing on how the cooler performs. 

Well in the box you get the dual tower setup with its two 120mm PWM fans already hinged with the metal locks on the heatsink. I really got to praise the fact that there were no bent pins on the heatsinks with the fans mounted on it. 

The fans claim to have a 69 CFM fan airflow and 2.19 mmH2O fan air pressure which sounds good for an air cooler considering there are 2 fans on the heatsink. 

The dimensions of the overall heat sink are quite clearly stated on its website. So no point in me reading numbers. 

There are six 6mm heat pipes passing through the thick copper base which has a slightly convex profile but really it’s so minute that I can’t really assess that properly. 

The AK620 weighs around 1.45 KG and is slightly heavier than the Noctua D15. But despite the more weight, the AK620 is considerably smaller than the D15 which might suit a lot of use cases. The other dual-tower option which is slightly shorter but wider than AK620 is Scythe Fuma 2 which isn’t available here in my country that easily. 

Well apart from all that, in the box you get a PWM splitter cable to connect both the fans to a single connector on the board. And their own thermal paste with a long L-type screwdriver. 

All the mounts and screws are metal and the backplate for Intel sockets is also metal. The finish on most of the stuff here is top-notch. 

These are the AMD and intel sockets the cooler supports. 

Overall from packaging to the cooler, the design aesthetics of the cooler are pretty calm and it doesn’t adopt any flashy elements over it.

The fins on the heatsink are structured in a way that they appear to have a different look at different angles so that could be something that was either done for either the looks or performance.

The top area of the cooler will appear almost completely black as the heatsink has these plastic covers on the top. So yeah I pretty much dig the design elements over the air cooler and the Deepcool logo has been seamlessly embossed on it.

The fans have a stealthy black matte design and nothing hurts my eyes here. The flashy heat sink totally gets hidden once you install the cooler inside a case. 

As I was going to test the cooler on a Ryzen 3900X, for the black plate you gotta use the stock backplate itself for any kind of AMD platform which the cooler supports. The 2-way screws are supposed to hold the backplate and the am4 mounts by themselves, with support of these other 4 screws which were inside the other packet – so don’t let that confuse you. 

I felt the cross-hatching of these 4 screws could’ve been better. It felt as if it might wear off quite easily if they were screwed – unscrewed a lot of times. 

But if you are going to install this air cooler and be done for a long long time then this shouldn’t bother you at all.

After applying the thermal paste I mount the cooler on the chip and place it right on top of the 2 mounting screws and then tighten them little by little alternatively. 

I had a long and comfortable screwdriver so this was quite easy. But if you don’t have one then the L-shaped screwdriver which came with the cooler would be sufficient to do this job as it easily cleared the height of the cooler. 

Because of the extremely logical design of the fans, I could rotate the fan and its clips however I wanted. Which helped me lead the fan wires directly towards the back of the case easily.

After that, I make sure the cooler tower has no weird free motion just in case. And then I proceed with installing the fan which I had removed earlier. 

After attaching the 2 PWM cables to the splitter, I plugged the connector into the CPU fan connector on the motherboard. 

And because I have 2 fans installed on the top of my case, it got a bit tough to plug the connector so it’ll be kinda wise if you plug in the splitter before putting the tower in.

You can see the sort of RAM clearance the tower provides against my 47.37mm tall XPG D40 RAMs. So yeah you might need to move the front fan a little upward to the extent your case permits if your RAM height is more than 47mm.

But if you ever want to alter or change or upgrade any of the RAMs in the future, removing just the front fan from the cooler would give you enough space to access all the RAM slots on the motherboard. 

All the tests were done inside the cooler master MB520 closed case. Which has 3 Silverstone fans in the front, 2 bequiet fans, and 1 stock fan up-top and in the rear. Yeah, it is the most confusing case ever. 

My primary cooler is the Noctua NH-U14S. So the AK620 will be compared against it. And I took a different approach in testing this time. The Noctua D15 performs better than the Noctua U14S which just very slightly performs better than the Noctua U12S. So just keep that in mind. 

I ran Cinebench R23 for all these sustained power load tests so each test ran for about 10 minutes. 

At a hundred watts the Ak620 performed quite a lot better than the U14S and the story was similar at 125 watts too. At 150 watts we notice a difference of about 2 degrees celsius. Which is about half the difference compared to the previous 2 tests. At 175 watts it ranged the same and at 200 watts the average temperature difference between the 2 dropped to less than 1 degrees. 

If we focus on the max temperatures, the thermal performance difference between these 2 coolers roughly matched the difference we saw during the average temperature charts. But at 200 watts load, both the coolers surpassed the 100 degrees C mark when the ambient temperatures were around 28.8 degrees. 

The Noctua U14S at the time of me buying it costed more than the Deepcool K620 and looking at how much better the Deepcool performed against it – we have a clear winner in terms of both pricing and performance for at least this specific comparison overall. 

Now because I do not have the Noctua D15 and if I had it, this comparison would have been even more fun. But there are these couple of reviews and they claim that the Deepcool does beat the D15 in terms of thermal performance. If that’s truly true then this air cooler has a frigging bright future ahead. But that’s a statement I really can’t vouch for till I do compare the 2 coolers myself. So yeah, there’s that. 

Even if we assume the AK620 performs just marginally worse than the D15 considering different setups can have slightly different results. Even then for the smaller footprint and the noticeably cheaper price – it would be pretty hard to ignore the Deepcool Ak620 as an option now. 

The noise tests will be in the video above at 6:57

🛒Amazon India:
🛒Amazon US: Awaited
🛒Amazon UK: Awaited

Hope you liked my efforts here. You can also hop onto our discord server ( for more chit-chat on relevant topics. Stay safe humans, that’s all for today. MuBot out.



Notify of
Newest Most Voted
Inline Feedbacks
View all comments
Roger Daigle

The Ninja 5 equipped with 2 XPG Vento Pro (Nidec GT) is a killer setup as well.
It’s even cheaper with cooler and the 2 XPG fans. My 3900x runs at 4300mhz@ 1.28v,
and gets 92c in FPU stress in AIDA ON CPU diode at 24c ambiant.