Review Deepcool Assassin III Review: Slaying the Competition

[Admin]

The Deepcool Assassin III goes big with a dual-tower, heatpipe CPU cooler aimed to manage temperatures on nearly recent platform aside from Threadripper. Skipping RGB lighting and relying on a pair of 140mm cooling fans, the Assassin III is a thermal force to be reckoned with.

Deepcool Assassin III Review: Slaying the Competition : Read more

 

[rickmeister]

The Noctua NH-D15 is The same price (Amazon) as the Assassin III. It has 140mm fans and seems to be a better heads up comparison to the Assassin III. Why wasn't it included in the test?

 

[msc545]

Why in god's name would someone choose a huge, ungainly air cooler over an AIO. The price difference is not that significant.

 

[Darkbreeze]

Probably because they don't have any NH-D15 test results on the same test platform that this test was conducted on. Since I believe they've created a newer, more standardized test platform at some point it is likely that there has not been a test of the D15 since doing so. I don't know for sure that that is the reason, but that would be my guess.

Obviously, a comparison of other twin finstack coolers would seem a lot more appropriate but not if the comparison cannot be done using the same test platform. That would just make it pointless.

 

[Darkbreeze]

Why in god's name would someone choose a huge, ungainly air cooler over an AIO. The price difference is not that significant.

There are MANY reasons why.

More expensive.

Doesn't particularly cool better unless you go with a very large AIO. 280mm is just about a requirement if you want to beat a twin finstack 140mm cooler.
AIO is louder. In some cases, MUCH louder, especially if you factor in pump noise.

Air coolers don't leak. And if you think AIO leakage is uncommon, you haven't been around them, or here, very often.

AIO coolers last, on average, about three years before pump failures force you to replace the whole cooler. Air coolers, you might have to replace a fan every six to ten years, maybe. If you keep the heatsink cleaned regularly and have intake filtration, you may not need to replace the fan for longer than that. And even if you do, 25 bucks for a good high end fan is a lot easier to swallow than another 100-160 bucks for a replacement AIO unit. In some cases, especially on cheaper AIO units, they pump may not even last that long. We've seen a number of units that barely made it out of warranty, or didn't even make it out of warranty.

An AIO is not like a custom open loop. It is not ever going to give you that kind of water cooled performance. AIO units main appeal is aesthetics, not pure performance. If you like the look, then there is much to be said for them. Otherwise, unless you go with a very good 280, 360 or 420mm cooler, you're not really going to gain anything in terms of performance over a big air cooler.

 

[msc545]

I disagree at least in part. AIO's are in the same price range (75-150) as air coolers, and adjusted properly, they cool much better than air coolers.

There is no way that an AIO pump is noisier than an air coolers fan(s). I've had AIO coolers going continuously for about 4 years now with no leaks or failures, although I admit a leak is something of major concern.

The best things about AIO coolers is 1) they fit in the case; 2) they are quiet); 3) they don't accumulate dust in the same way air coolers do and are easier to clean; 4) they cool better; 5) they don't weigh so much as to distort sockets and motherboards, which some air coolers do; 6) you don't have to remove an AIO to transport a computer in the upright position.

 

[Virtual_Singularity]

Why in god's name would someone choose a huge, ungainly air cooler over an AIO. The price difference is not that significant.

Risk of leakage, corrosion, pump probs, etc inherent with AIO water coolers & even costly closed loops, only grow with time. That some water AIOs are beaten by air coolers, or the cost of some AIO's being expensive vs air coolers for negligible results is also weighed (sry for pun) by many consumers when making their choice.

The Noctua NH-D15 is The same price (Amazon) as the Assassin III. It has 140mm fans and seems to be a better heads up comparison to the Assassin III. Why wasn't it included in the test?

Valid question for sure, but not surprised at all the D-15 wasn't included here. I'd bet the Deepcool is likely
better or similar to the stock D-15, and in another test chart from another site it showed just that, though not by much. I know this prolly won't be too helpful, but fwiw my experience showed the newer, more compact NH-U12A slightly beating/on par with the D-15. Tho depending on processor type, size, load under OC, ambient temps, and many other variables, it's really always a toss up in re to charts, when it comes to temps most of all.

 

[Darkbreeze]

There is no way that an AIO pump is noisier than an air coolers fan(s).

Nobody said the pump alone was noisier. You want to run your AIO with just the pump? I didn't think so.

The noise levels for practically any 240-420mm AIO is always going to be louder if you factor in not only the two fans that will be running but the addition of pump noise as well. Even coolers with two fans on the heatsink are typically quieter because unlike an AIO they are buried inside the case rather than being faced directly against an opening to the outside. It's really no different that taking an alarm clock that is ringing and putting it in a box. Sure, you can still hear it, but it's going to be quieter than it was before.

This is obviously not directly comparable but it is in fact representative of what you'll find, across the board, doing similar comparisions. Note that practically every AIO except the Floe Ring 360 is louder than the nine air coolers in the comparison. And the Floe Ring is a 360mm so clearly it can run it's fans a bit slower while still maintaining fairly decent performance since it has a lot more surface area AND an additional fan.

Overall, AIO coolers are much noisier, with 50-60db not being uncommon under load. Sure, you can tune them for quieter operation but you are going to lose a significant amount of performance that way. By comparison, The NH-U14S or NH-D15 can be run at full speed fans and with a decent case the noise levels are low enough to be conversational. Spin it however you like, complain about the full speed 12v operation, the fact remains that most AIO coolers are going to have significantly louder cumulative noise levels with it's two fans and pump than an air cooler is. At least if it is a decent quality cooler with good quality fans.

And I don't ever have to check to see if the damn thing has ruined my hardware today or not. Nor do I have to guess if my thermal issues are due to a bad pump. Nor will it cost me another 120 bucks when something does go wrong with my cooler.

 

[PapaCrazy]

Why in god's name would someone choose a huge, ungainly air cooler over an AIO. The price difference is not that significant.

Everyone's raised good points. I'd just like to point out the risk for single point catastrophic failure is much higher due to increased complexity.

Single point of failure worst case scenario air cooled: One of the two push-pull fans blows a bearing. You chug along with one fan at higher temps. CPU throttles slightly. Maybe you get really unlucky and even the second fan breaks. You still have a large heat sink. It cannot fail. Redundancy in cooling. More throttling, but CPU chugs along. Fans can be replaced same day without removal of entire system.

Single point of failure worst case scenario water cooled: Pump breaks, and water sits still over the CPU block like a thermal blanket. CPU temps continue to increase, despite attempts to throttle. TJ Maxx of 100c is boiling point for water. Fluid near the CPU block cannot distribute its heat, climbing toward 100c and increasing pressure inside the tubes. Pressure causes tubes to rupture, hosing down motherboard and GPU. Things suck for a while.

 

[rickmeister]

Valid question for sure, but not surprised at all the D-15 wasn't included here. I'd bet the Deepcool is likely
better or similar to the stock D-15, and in another test chart from another site it showed just that, though not by much. I know this prolly won't be too helpful, but fwiw my experience showed the newer, more compact NH-U12A slightly beating/on par with the D-15. Tho depending on processor type, size, load under OC, ambient temps, and many other variables, it's really always a toss up in re to charts, when it comes to temps most of all.

The U12A may keep up at lower TDPs but my 8700K ran 12C cooler with the D-15 over the U12A and it took longer to saturate the cooler. Recovery time was better on the U12A.

 

[rubix_1011]

The NH-D15 has not been tested on this platform; DarkBreeze is correct. It's an older cooler and I haven't been sent one from Noctua to test. The be quiet! Dark Rock Pro 4 is one of the largest/newer coolers I've covered and have normalized test data for. Since I took over the cooler reviews a few years ago, the platform needed to be re-benched and we needed a new place to start. I apologize for not having that data for comparison, but allow me to explain.

The NH-D15, while being excellent for its age, was likely covered before the period when I had the test bench setup. I don't have access to every cooler available - I only have test data on coolers which have been supplied by vendors to be evaluated and get coverage on the site. I have actually asked Noctua for an NH-D15 sample that I can test and use for reference and while new models are often the priority, sometimes getting older components takes longer to get responses - hopefully I'll get one soon, I really would like to do a larger round-up article on large air coolers vs. AIOs, so it's rather ironic this conversation is taking place here.

To address other points of big air vs AIOs: Honestly, I'd personally rather have a larger air cooler than an AIO at the same price, and I'm a custom watercooling guy of more than 18 years. I do have some newer AIOs that are in my queue to be evaluated that are incredibly good looking, though.

While AIOs have gotten far better in recent times, quality air cooling has also come a VERY long way in the same period while also still remaining immune to the biggest fear most PC tech forums face: 'leaks'.

Most big air coolers are also nearly silent, or at least quieter than closed loop liquid cooling. The same cannot be said for a 240mm AIO...go compare the dbA on my test comparisons.

 

[Phaaze88]

The U12A may keep up at lower TDPs but my 8700K ran 12C cooler with the D-15 over the U12A and it took longer to saturate the cooler. Recovery time was better on the U12A.

And this is one of the problems with some of today's cooler reviews. If the cpu cooler isn't actually being pushed, one doesn't really see what the cooler is actually capable of.
Testing those 2 coolers on cpus like a 7700K or a Ryzen 3700X is seriously flawed - if it were something like a 9900K instead, the results wouldn't be as close, I'd think.

"... it took longer to saturate the NH-D15. Recovery time was better on the U12A".
I believe that has to do with the push-pull orientation on the single tower being more effective VS the push-pull on the dual tower.

 

[mdd1963]

The Deepcool Assassin III goes big with a dual-tower, heatpipe CPU cooler aimed to manage temperatures on nearly recent platform aside from Threadripper. Skipping RGB lighting and relying on a pair of 140mm cooling fans, the Assassin III is a thermal force to be reckoned with.

Deepcool Assassin III Review: Slaying the Competition : Read more

Someone compared the Deepcool Assassin III to an NH-D15 and they were both within 1C on temps under load....

(Another website said the Deepcool was quieter, which , considering how quiet the D15 is, would be amazing...

 

[Phaaze88]

Someone compared the Deepcool Assassin III to an NH-D15 and they were both within 1C on temps under load....

One of these?

Deepcool Assassin III

Attempting to dethrone Noctua's NH-D15.

hexus.net

8700K up to 4.8ghz in a Dark Base 700

Deepcool Assassin III Air Cooler Review - KitGuru

Today, we have a new high-end air cooler for review. The Deepcool Assassin III was first seen at CES

www.kitguru.net

9900K up to 4.9ghz, open test bed, and in a Deepcool Macube 550

https://www.guru3d.com/articles_pages/deepcool_deepcool_assassin_iii_review,1.html
4790K up to 4.6ghz, open test bed

 

[drivinfast247]

Nobody said the pump alone was noisier. You want to run your AIO with just the pump? I didn't think so.

The noise levels for practically any 240-420mm AIO is always going to be louder if you factor in not only the two fans that will be running but the addition of pump noise as well. Even coolers with two fans on the heatsink are typically quieter because unlike an AIO they are buried inside the case rather than being faced directly against an opening to the outside. It's really no different that taking an alarm clock that is ringing and putting it in a box. Sure, you can still hear it, but it's going to be quieter than it was before.

This is obviously not directly comparable but it is in fact representative of what you'll find, across the board, doing similar comparisions. Note that practically every AIO except the Floe Ring 360 is louder than the nine air coolers in the comparison. And the Floe Ring is a 360mm so clearly it can run it's fans a bit slower while still maintaining fairly decent performance since it has a lot more surface area AND an additional fan.

Overall, AIO coolers are much noisier, with 50-60db not being uncommon under load. Sure, you can tune them for quieter operation but you are going to lose a significant amount of performance that way. By comparison, The NH-U14S or NH-D15 can be run at full speed fans and with a decent case the noise levels are low enough to be conversational. Spin it however you like, complain about the full speed 12v operation, the fact remains that most AIO coolers are going to have significantly louder cumulative noise levels with it's two fans and pump than an air cooler is. At least if it is a decent quality cooler with good quality fans.

And I don't ever have to check to see if the damn thing has ruined my hardware today or not. Nor do I have to guess if my thermal issues are due to a bad pump. Nor will it cost me another 120 bucks when something does go wrong with my cooler.

That chart is useless without context. What were CPU temps. at 100% fan speed?

 

[Darkbreeze]

No it isn't. Don't be obtuse. But we'll put it up, so you can have your context and eat it too. Which, by the way, the Noctua wins there too. So, quieter and cools slightly better. Yay, another win for air. Several of them in fact. Jeez, look at the Le Grand Macho RT. It's quieter than the D15 AND outcools it. At least on THAT particular test bench. If you don't like the results, just find another review. Assuredly the results will be different.

 

[rubix_1011]

My reviews utilize an i7 5930k which is a 140w TDP part at factory clock speeds (3.50Ghz); I have it at 4.2Ghz @ 1.20v across all 6 cores. I have actually had some coolers wilt trying to keep this CPU cool - those with flawed design or manufacturing defects .

The same exact CPU, vcore. clocks, motherboard, RAM, SSD, thermal compound, case fans are used for every test; the only difference being the cooler being evaluated. Eliminating variables allows direct comparisons to be made over time without needing to continually retest every cooler.

 

[rickmeister]

Someone compared the Deepcool Assassin III to an NH-D15 and they were both within 1C on temps under load....

(Another website said the Deepcool was quieter, which , considering how quiet the D15 is, would be amazing...

As would be expected, the Assassin III is a clone of the D15. The only problem I see is that when you price a clone product the same as the original the buyer has no incentive to move to your "untested" product. If it were priced in the $60 range and performed equally as well as the D15 then you have a compelling reason for the consumer to switch.

 

[Darkbreeze]

My reviews utilize an i7 5930k which is a 140w TDP part at factory clock speeds (3.50Ghz); I have it at 4.2Ghz @ 1.20v across all 6 cores. I have actually had some coolers wilt trying to keep this CPU cool - those with flawed design or manufacturing defects .

The same exact CPU, vcore. clocks, motherboard, RAM, SSD, thermal compound, case fans are used for every test; the only difference being the cooler being evaluated. Eliminating variables allows direct comparisons to be made over time without needing to continually retest every cooler.

Yeah, I wasn't trying to bash your reviews, or results. I was simply saying that for any given cooler, if you don't like the results on one site, go look on another site, because you're likely to find it higher or lower than the results against the same coolers on another site. I have no lack of faith in your review process.

 

[rubix_1011]

Oh, no, not at all...I'm open to criticism and feedback and most of my response actually wasn't directed to you, mostly as the group discussion as a whole (just to try and clarify my testing methods and setup).

I often look at different reviews on coolers that I'm covering as well (assuming it isn't embargoed) to see how it is being tested. Many sites use 'popular CPUs' which may or may not thoroughly load a cooler as much as notoriously 'hot' CPUs. If we were to re-vamp the CPU large cooler testing bench, I would say the 9900k is a good chip to use, much like the 2990WX Threadripper CPU I use for TR4 tests.

The difficult part is that different hardware is released on regular cadence which means that if we continuously change the hardware platform, it invalidates all previous tests on previous hardware. Maintaining some form of continuity allows for 1:1 cooler comparisons, provided the hardware is adequately loading the coolers in the same manner, each time.

 

[hotaru251]

i mean....why would I pay $89 for this cooler when its barely less than a nh-d15 which is basically the best air cooler on market and beats/rivals aoi's?

also noctua's support is amazing (which i would alone value worth the extra $)

( chroma is actually visually pretty too)


i do question why you used a single tower (12a) to compare to a dual tower.

 

[Darkbreeze]

1. Noctua isn't available everywhere.

2. Noctua is MUCH more expensive than Deepcool in SOME regions.

3. Some people SERIOUSLY hate the Noctua earth tone colors, and the black versions are about twenty bucks more expensive than the standard Noctua colors, which is kind of dumb since I'm pretty sure that manufacturing with the black materials isn't any more expensive than manufacturing with brown or tan. Maybe slightly more expensive for the black coating on the heatsink but honestly, I wish they'd just not do that and include an option for a black fan on the chrome heatsinks anyhow. Having the option for blackout style heatsinks is fine, but a lot of us would prefer to simply have the black Chromax fans on a chrome heatsink.

4. The D15 ISN'T the best heatsink on the market, the D14 is, as evidenced by this rather thorough review.

Noctua NH-D14 Redux: Modern Fans On An Old Heatsink - Overclockers

Various combinations of fans were tried on the NH-D14, especially the NF-A15 and the NF-A14. We found a winner.

www.overclockers.com

Unless of course you believe the review done at Kitguru on the Zalman twin finstack cooler, which I have serious doubts about as evidenced by the comments I made regarding some discrepancies I found in the testing procedures across the site reviews.

Zalman CNPS20X and CNPS17X: Noctua didn’t see this coming! - KitGuru

Since publishing our review of Zalman's new CNPS20X and CNPS17X air coolers, we have received a numb

www.kitguru.net

i do question why you used a single tower (12a) to compare to a dual tower.

He already answered this question, as did I, when it was explained that the coolers which have been tested on the test platform, which is relatively new, are what you must use for comparison of any other coolers on the same platform. It's not like these guys generally get these parts to review and then they are just hanging around forever for them to review again and again indefinitely each time the review test bed is changed. Many of them must be returned to the company. Usually only the ones that have been purchased by the review site or the reviewer remain on hand for a revised test on the new test platform. I'm sure if you'd like to purchase a D15 and send it to him he'd be glad to test it on the new platform and include the results in future comparisons.

 

[Joe02a]

One Confused question, I see some sources state 171.5 and others as 165 is height? I like the ram clearance and look. That is what has my interest. My case is the NZXT Source 530, so not worried on the height. Just confused.

 

[Phaaze88]

One Confused question, I see some sources state 171.5 and others as 165 is height? I like the ram clearance and look. That is what has my interest. My case is the NZXT Source 530, so not worried on the height. Just confused.

Part of that confusion is Deepcool's fault. See here: http://www.gamerstorm.com/product/CPUAIRCOOLER/2019-05/1287_11125.shtml
They posted official measurements of the cooler's height without taking into account memory clearance for the front fan!
If you look up images of it, the front fan HAS to sit on top of the memory.

 

[sean8102]

Is Deepcool a unknown brand among PC enthusiasts? I know Noctua is VERY well known but it's not like Deepcool is some fly by night brand or anything.

But anyways I'm done with AIO's personally. My reason? Bought a H100i V2 late 2017 and it died October 2020. The pump just completely died out of nowhere. With air coolers I KNOW it's gonna last a LONG time. Worst case one of the fans messes up. I can just put a new one. I already tried the Noctua NF-F12 PWM fans from my h100i v2 (because the 1st problem I had with that AIO was one of the fans making a horrible buzzing sound when I got it) and those notua fans clip right on perfectly. Also, the cooling temps with my H100i V2 were good but never that great, and not as good as the Assassin III. With the h100i v2 stress tests would see cores peak at 78c. On the same OC I have now (4.8 Ghz all core - 1.230v). Same exact stress test with the Assassin III and none of the cores peak above 75c and even when I try to listen for it, I can't hear the Assassin III. It's silent.