Deepcool Gammaxx 400 Slim-Tower CPU Cooler Review

[Crashman]

Deepcool Global pushes its value message with low pricing, but can its $30 Gammaxx 400 provide enough cooling power for a high-end processor? We compare it to two of our favorites to find out!

Deepcool Gammaxx 400 Slim-Tower CPU Cooler Review : Read more

 

[Traciatim]

Personally, I think this would have been a better comparison if it was done with the current value champ of forum recommendations, the Coolermaster 212 EVO. That has always been the go to cooler for recommending to people with new builds or not very specific requirements because it's fairly cheap and works well in almost any case.

 

[kunstderfugue]

I'd have loved to see this cooler compared to the 212 EVO. Because if it's as good as these tests show it to be with that price tag, we might have ourselves a new King of the Hill.

 

[Val Cyril Estrada]

I dont know why people are spending almost a hundred dollars on a noctua when you can get this for the same or better cooling performance. Im using this but replaces the LED fans with Corsair SP120s since i have black and white themed system with deepcool color changing RGB lights

 

[Onus]

I really hope this finally silences the incessant parroting of the [overpriced] Hyper212 EVO. I'd seen a good review of the 300, and I'm glad the 400 is consistent with that.
Nice job, covering a budget cooler!

 

[Calculatron]

Wow, I am seriously impressed by this.

 

[Chris Droste]

who pays $100 for the Noctura? that's a $65 part all day, Val.

 

[Onus]

Thomas, you have batted one out of the park with this one. Imho, solid, budget parts are something people want to see, in any / all categories.

 

[zodiacfml]

I reckon, this seems to perform slightly better than a Hyper212 EVO.

 

[Darkbreeze]

Great review, couple of things I feel I need to interject though. As usual, and as mentioned in the Shadow rock slim review comments, not offering comparisons to the 212 EVO, as mentioned here already, makes little sense when you're trying to show budget cooler performance and value.

The 212 EVO is the budget cooler that 90% of recommendations focus on, therefore, you need a review of that unit and you need to begin offering comparisons to it.

Secondly, I'd REALLY like to see you retest this unit with a higher quality fan installed. If it's as good as your results indicated, then I think the addition of a higher quality bearing fan like the NF-F12 PWM or one of the Cryorig, Thermalright, Scythe or Phanteks fans that have quieter, longer life bearings would be a great thing to look at since it would still keep the price under fifty bucks.

If the performance is still good with one of those fans, it makes it an even better option versus something like the U12S at 65.00 or more.

 

[Xaltar]

I have been a fan of DeepCool's products since I picked up a Gammax 300 a while back. While their fans leave a lot to be desired their tower coolers offer fantastic performance for the money. My Gammax 300 even cools my 6600k respectably when I am switching it in and out of review boards (simple mounting system makes it handy for in and out jobs). I still prefer my Thermaltake Frio for hardcore overclocking but for a 24/7 light OC the gammax does the job well, that and I don't have to worry about warping my CPU substrate.

 

[10tacle]

The Cryorig H7 dethroned the former king of entry level/budget air coolers, the 212, last year. It's priced in the same category (no greater than $35US) yet beats it by several degrees. Why it is still the most widely lauded/mentioned budget cooler is a mystery to me.

 

[Darkbreeze]

It's no mystery. The EVO is usually about 8 to ten bucks cheaper, and people are cheap. Plus, Cryorig has shown that they cannot meet the supply with the demand. There have been extended periods when the H5 and H7 have simply not been available anywhere in the world.

 

[Co BIY]

"We can also be relatively certain that the fan bearings won't last as long, since Deepcool warranties the unit for a mere two years."

I think this statement deserves follow-up. I'm not sure warranty length is always an indicator of quality, sometimes it's an indicator of how much time and effort a company is willing to expend on the most difficult 5% of consumers who will always maximize their warranty returns by fraud. Sometimes it's an indicator of the quality signal they want to send to the market independent of actual product quaility.

What kind of bearings do each of these fans have ? What are the advantages of each ? When my "cheap" OEM fan burns out what should I replace it with for improved longevity/performance ?

I think for many people having to do any maintenance, even as simple as fan replacement, in under four years would negate any saving gained by a "budget" cooler. But I really see no support in the article for the contention that the GAMMAX 400 fan is of lower quality than the compared units.

 

[technoholic]

Some coolers i'd have liked to see in this review would be: Cryorig H7, Silverstone Argon AR01, Xigmatek Dark Knight and of course CM hyper 212 evo. This way it could be a nice reference article for budget/average builders

 

[Co BIY]

I know most review sites are dependent on manufacturers to send samples for review purposes but for lower margin/budget products sometimes a consumer provided sample is the only way to get a review (There is also value in market purchased samples vs. handpicked manufacturers samples). Maybe someone out there is interested in providing Thomas a Coolermaster 212 EVO for testing.

 

[Darkbreeze]

He has already said in the past that he has one in the pile and just needs to find time to review it. I think maybe that ought to be a priority. Not necessarily for the purpose of it's own article, but to further increase the relevance of other articles, since he will then have those results to draw from for comparison.

 

[alidan]

honestly, what needs to happen is just one month, benchmark every single cooler you have available with whatever base system you use (be it amd or intel) with its stock cooler then oc it to as high as the stock will allow, and that is the base that all other coolers are judged by.

then, oc it again with the best cooler, making sure you have a damn good oc chip (4.5+ if i remember some of the 4000 line can hit 4.8 or 4.9) and work backwards till you can no longer handle the oc on a cooler.

then oc the chip (if it needs it) to 4ghz and again, top down till nothing will handle it.

once you do that, replace the stock cooler fans with whatever the hell the best noise/db ratio fans you have are, and run them 100%.

personally i chose bgears 120mm with 100+cfm as my cooler for my case, sadly with stock cooler as i have no back plate access and i did not want to rip my motherboard out and risk damage when its not needed.

that way you have several factors to go into play.

1) how much better then stock is the cooler,
2) how much of an oc can this cooler handle
3) an oc tier list, as in these can handle moderate, these can handle extreme, these just offer better/quieter over stock
4) how good is the cooler with stock fans
5) how good the cooler is if you disregard stock fans.

it may be a pain to do the initial benchmark but you would effectively make a comprehensive list that tells you everything you could possibly need to know about the coolers.

 

[Darkbreeze]

^^Yeah right. This is a JOB. These guys are working. They have specific criteria and topics mandated by the editorial management that have to be done if they want to get paid. I really doubt having the time to do what you've suggested is ever going to be an option unless they are willing to do it on their own time, without pay. I'd settle for just doing the one cooler that gets referenced more than any other, which IS the 212 EVO.

That way there is a well known baseline for other coolers to be compared to, and we can point to something that shows exactly why the EVO is not the value many think it is. And that it's loud as hell under a full load with anything bigger than an i3 or A series APU.

 

[Co BIY]

I wonder if there would be a way to make a cooler testing rig that replaced a full blown reference computer system with a "tiny hot plate" that could be made to heat at different reference loads. That way you wouldn't need to to build a computer each time you ran one of these tests. That way you could see how each cooler performed at each "load" level.

The only issue I see with this is that fan controls might be difficult to control but I think the advantages would outweigh the challenges at least over multiple tests. I'm sure the manufacturers do not test during development using actual CPUs running on regular motherboards.

 

[Darkbreeze]

A hot plate would not have the same characteristics as any given CPU, much less fifty different CPUs. You can test the output of an engine on a dyno, but until you put it in the chassis you want to run it in, you are just guessing how it will perform.

 

[10tacle]

It's no mystery. The EVO is usually about 8 to ten bucks cheaper, and people are cheap. Plus, Cryorig has shown that they cannot meet the supply with the demand. There have been extended periods when the H5 and H7 have simply not been available anywhere in the world.

The cooler is 14 months old now on the market. Is it still that hard to find on occasion? If so, what a manufacturing logistical fail by Cryorig. I have not been shopping for one, so I have no credibility to question anything.

With that said, and this is just my opinion, if I was looking into building a new PC and having to nickle and dime worry about spending $10 more on a certain PC component that offered a better performance value (especially when wanting to mildly overclock), I'd be reassessing my finances and questioning the wisdom of building a PC altogether. But, that's just me.

 

[Darkbreeze]

I agree, but unfortunately, that IS the case. Hundreds and hundreds of people, on a weekly basis, want to built bazillion gigsmoosh systems that run in the thousands of dollars, and then skimp and refuse to pay more than 50 bucks on a power supply or cooler. We can't fix stupid, all we can do is make suggestions.

As far as the Cryorig supply issue, yes, it just happened not two weeks ago, again. Nobody had the H5 or H7 in stock for almost two weeks. In fact, the Cryorig rep posted a comment in one of the threads I was participating in to say they were aware of the shortfall and were working overtime to get them back in stock. Cryorig is a rather small company, all things considered, and it's not surprising they can't keep up when the demand for a very good product is higher than expected. They don't have the facilities of companies like Corsair or Cooler Master.

 

[Co BIY]

A hot plate would not have the same characteristics as any given CPU, much less fifty different CPUs.

I disagree. I think that a controllable hotplate could much more easily be used to emulate the important characteristics (heat output) of many different processors in many different work states (idle, taxed, OC'd) and without ever removing the cooler being tested. Just change the hotplate's input current to emulate a max load of 35 TDW for a low power HTPC chip or 130 TDW for an OC'd i7-4790K.

I think the engine/dyno example is very good comparison. If I want to test engine-to-engine then putting them both on the same dyno (especially if my dyno can consistently emulate different loads) is better than testing them on the road in different cars. The reviewer in this article was careful to test the new cooler (engine) in the same car (cpu/mb) on the same course (steady taxing workload on CPU).

The biggest variable that was not controlled for or at least not discussed was ambient room temperature for the various tests. Especially when comparing results run weeks or months ago for an air cooled system.

For an appropriately small hot plate a good source might be an e-cigarette vaporizer plate or similar.

 

[Darkbreeze]

They test at the same ambient, in a controlled room, so ambient is a non-factor for Crashman's tests.

I'm not sure how you think you would "emulate" any specific load to match what MIGHT be seen on a given CPU.

Watts? Wouldn't work. Running the same amount of wattage through a hot plate is not going to give you similar heat results as seen from a CPU at the same wattage.

Figure out how many watts causes the hotplate to heat to the same temp as the CPU at a given wattage or load? Wouldn't work. The hotplate would never dissipate or exchange heat at the same rate as a CPU, so it would be inaccurate.

Plus, any hotplate would have further different heat exchange characteristics than a CPU because the physical characteristics of the hotplate wouldn't be the same as the CPU lid, and the hotplate would not be locked down into and onto a substrate. Too many variables in my opinion. I'm sure if this was viable, it would have already been done.

Airflow around the hotplate would change things as well, since the CPU would be unlikely to have the same volume of air AROUND it, being so close to the motherboard, which would also affect the results as heat is exchanged to the air as well, which is why fanless heatsinks work well. The hotplate would have more exposed surface area than a CPU would, by far.