Hey guys,

I'm debating these two coolers and mainly basing my choice on the SPCR reviews and price. I am however a bit confused by the numbers of the Mugen 2 presented by SPCR.
On page 6 in the review itself (http://www.silentpcreview.com/article961-page6.html) it gives the following numbers for a single nexus reference fan:

12V @ 16 dBA - 36°C
9V @ 13 dBA - 38°C
7V @ 12 dBA - 40°C

However in the comparison chart seen in reviews for other heatsinks (as seen here: http://www.silentpcreview.com/article1039-page9.html) it gives the following:

12V @ 16 dBA - 39°C
9V @ 13 dBA - 42°C
7V @ 12 dBA - 45°C

Why this difference in temperature and which data should I be looking at?

According to the chart, the NH-D14 has these stats:

12V @ 17 dBA - 38°C
9V @ 14 dBA - 41°C
7V @ 12 dBA - 43°C

Only a 1° difference between Mugen 2 and the NH-D14 at the expense of a 1dBA increase, except for 7V.

Am I right in assuming that in this comparison the Mugen 2 is running with only one fan and that the NH-D14 is running with its normal dual fan configurations?
If so, I assume adding a second fan to the Mugen 2 will make it cooler and quieter than the NH-D14?

The NH-D14 is $129 where I live (currency converted to USD) and the Mugen 2 is $68. However, buying a second slipstream 1200rpm fan, fan clips and a MX-3 thermal paste brings the total cost to to the ideal Mugen 2 setup to $109. The Noctua cooler seems to have good fans included and thermal paste so I would not add anything to that, also I find it very expensive as it is.

I would get the NH-D14 if it cools better than a dual fan Mugen 2 setup as the price difference is not that high, however if the Mugen 2 with two fans is a better performer then it's a no-brainer.
I am confused by these results by SPCR though, and also that reviews from other sites suggest that the NH-D14 is a considerably better cooler than the Mugen 2 in comparison to SPCR.

These reviews are not that accurate. Not from site to site, or even test to test at the same website.

They can't be. You have sample variances, measuring variances, different mounts (some better than others). The world just isn't that tidy.

But these reviews do permit you to stratify and rank different cpu coolers.

The two you are looking at are both good. You will not make a mistake selecting either one.

A third one you should consider is the Noctua NH-C14 Dual Fan Top-down CPU Cooler
http://www.silentpcreview.com/Noctua_NH-C14_CPU_Cooler
It will do a better job of keeping your motherboard components cool.

The test were done on two different test platforms, as lined out on page 4 in each review.
One is the old CPU platform, featuring a historic Pentium 4, the other is SPCRs current platform with a Nehalem based Core i7.

The old test even features a seperate entry for the Mugen with 2 fans. Both data is relevant. You can cross-compare the tables. With 2 (Nexus) fans at 7V, the Mugen bested the Megahalems by 4c. On the new test bench, the Megahalems beats the Noctua NH-D14 by 1c. Doing the math: the Mugen should have beat the NH-D14 by 5c.

BTW: every CPU heatsink I've ever purchased came with thermal paste.

What are you trying too cool? and it what case/airflow?

Is it 1366 i7 OC that you need a Mugen 2 or Noctua NH14D in push+pull? How much heat do you need to get rid of? If its anything lower than 130W TDP you should rethink the dual fan setup IMO.

Mugen 2 with 1 fan can cool most CPUs and do so quietly. That would have been my choice given the price difference. I thinks Mugen 2 includes Silmore thermal paste (it sucks!!). AC MX 2/3/4 is much much better.

If you do decide to go with 2 fans on the Mugen, 1200RPM is too high. The bundled Slipstream PWM is rated at 250-1300 RMP and isnt gonna get to more than 500-600 on your typical motherboard control. You want the second fan to slower to start with, and the ability to under-volt it even more to 500-600 range. Maybe a Slipstream 800, or S-flex 800. 2 matching Nexus 120s would be ideal.

Between D14 and Mugen, I'd go for the D14 because it has better mounting, better overall build quality. Hard to find a performance comparison of D14 vs Mugen in dual fan configurations. For what it is worth though, the D14 is significantly better than Ninja 3 when both are using two Scythe KamaFlex fans ( http://www.dexgo.com/index.php?site=artikel/view.php&id=535&rubrik=Hardware&seite=7 )

From the same review, you can see that the Thermalright Archon is the new single fan / low speed king ( http://www.dexgo.com/index.php?site=artikel/view.php&id=535&rubrik=Hardware&seite=6 and http://www.dexgo.com/index.php?site=artikel/view.php&id=535&rubrik=Hardware&seite=8 ). It also performs very well with dual fans... so I'd recommend getting the Archon if it is available.

My last three coolers are Megahalems, D14, and Silver Arrow (still in use). I would recommend Megahalems for normal use, and Silver Arrow for overclocking.

Some interesting opinions here, much appreciated.

My CPU is an AMD Athlon II x6 1090T 3.2Ghz, 125W TDP and a P183 case. I'm interested in trying some overclocking although it will be my first, so I'm not sure if I'll be doing much of it or not. If the cooler is not up to the task of running it within temperature limits on an OC and still be quiet I'll probably not be overclocking it much.

Yes the Mugen 2 comes with thermal paste, but I've yet to hear anyone say something nice about it. Thus I didn't consider buying a Mugen without spending money on a better paste.

Interesting but also disappointing, I was hoping to just add another Slipstream 1200rpm as it is the only SPCR recommended fan available in my country and are reasonably priced. The Slipstream 800rpm, S-Flex and the Nexus are not available here.

As for the alternative coolers suggested here, unfortunately none of those are available to me locally.

I think maybe getting the Mugen 2 and MX-3/4 paste would be a smart choice. It's far more economical and judging by your feedback it may just be enough for me. I can then add to it if I feel it's not enough and still end up under or around the D14 price it seems.

The Slipstream 1200rpm is a good fan. You just need to undervolt it.... in fact it really isn't that bad even at 12 volts. My recollection is that it is easily undervolted to 800 rpm - making it basically the same at the 800rpm slipstream operating at 12 volts.

And the Mugen will perform just fine, regardless of the mounting orientation.

I have a Mugen 2 in my test bench on a hexacore AMD rated at 125w TDP.

With Cool 'n' quiet mode disabled, it constantly runs at 2.8 Ghz and the stock fan keeps it at 45c with an average load of 1.5% per core at any given second. The fan is not audible at that setting.

Alright, I'm gonna go for the Mugen 2 and try it out with just one fan to begin with. I have another slipstream in my case that I could use to try it in push-pull and see how I like it.

Thanks everyone!

Ok a newb question here... What's the best solution for adding a second fan to the Mugen 2 if the second fan is only a 3 pin fan? The included fan is a 4-pin 1200rpm slipstream with PWM control, however as far as I can tell you can't buy another one seperately. Is there a solution for that without having the second fan running at full speed or getting a fan controller?

If I already had another Slipstream in the case, I definitely would not get another fan. I also would not waste $ on other thermal paste. Just get the Mugen, nothing else. Chances are, you will figure out how to put your 2nd fan on the Mugen and control its speed if needed. I have setups where there are only 2 fan in the case -- in push/pull on the heatsink... and since the fans are in direct line with the 120mm exhaust hole on the back panel, no case exhaust fan is needed.

1. I would just follow Mike's advice.
2. You actually can buy a second scythe PWM fan online like the one that comes with the Mugen. (I would try first at frozencpu or performancepcs)
3. It is unlikely that you will need a second fan. The Mugen has as more capacity to cool, than the amount of load you are likely in real life to be throwing at your CPU... unless your primary use of your computer is to run benchmarks. My recollection is that you have a 6 cpu AMD. It will be difficult for you to ever push it past 15 or 20% utilization.
4. After you install the Mugen, if you really want to see what a second fan will do, just hook up the Scythe you already have as a second fan and see what happens running it at full speed. Then you can decide if it is worth spending money on extra fans or not. No matter what you do, you can only get so close to ambient temperature. Physics imposes limits as to how close you can get by turning up the wind. Those last few degrees are killers and just aren't worth it.

FYI see:
http://www.frostytech.com/articleview.c ... cleID=2259
(an 87.4dBA 6000 rpm military spec fan on a Scythe Orochi heatsink)

Thanks for your opinions, sounds like I'm going for an overkill here. I'm a bit surprised though, although I must admit lacking experience.

15-20% utilization? I play a lot of games on my computer and I'm pretty sure some of the cores go to at least 50% usage in games or even higher. Also, I have a Radeon 6970 in the case (P183) and this thing gets hot, over 90° in some games. Are you guys sure that there is plenty cooling going on under these conditions without the single slipstream fan on the CPU cooler being overly noisy? Or maybe it would be running at full speed even if I had another fan?

I understand it can only be cooled so much, but my concern is what level of RPM/dBA is needed to achieve that.
Well, I'm ordering the cooler as is as advised and perhaps that will be all I need.

2: According to SPCR's review it's this model. http://www.scythe-usa.com/product/acc/0 ... etail.html
3: I have a hexacore AMD, never have to worry about it causing lag.
4: I bought another SY1225SL12LM-P and fan clips. (You'll want type A) I bought it because I couldn't feel air moving past the back of the cooler, now I have 2 fans and can barely feel the breeze back there. Temps are still in the high 30s.

The radeon is a separate issue. It will have its own cooling issues and noise issues. You will just have to see what you find and deal with that then.

Though a downdraft CPU cooler like the Noctua is more likely to help with the video card's cooling than a tower cooler like the Mugen, even that is unpredictable.



Noise levels are very subjective. I would expect your card to be the source of most of your noise.

The Mugen is probably one of the top 5 or 6 cpu air coolers that exist today. You may be pushing the video card and you may be pushing 2 of your 6 cores at those times you are gaming. It is still quite possible (in my opinion likely) that your CPU cooler fan will be able to handle it all while loping along at a quiet 700 rpm. And if it can't it probably means it is your case that needs more air circulation, not your heat sink. All you should expect from the heat sink is to move heat from your CPU to the air inside your case.

Still you never know until you try. Your video card should be expected to generate a fair amount of noise and heat.

But even at 50% utilization of 2 of your 6 cores - that still works out to a little over 15% utilization. And if your video card is cooking your case, the fan you need is a fan to remove hot air from your case, not an additional fan for your CPU heat sink.

Bottom line: Just follow Mike's advice. Expect your noise and heat issues not from the heat sink, but from the video card. I just don't see the mugen being your problem.

What are your average and/or typical CPU utilization percentage(s)

(a) When you were running a single fan, what were your temps.
(b) How much did your cpu temps go down when you added a second fan?
(c) What speed did you run the single fan and speed(s) did you run when you added the second fan?

That changes from setup to setup. In my experience the best way to keep temperatures down in conventional cases without being audible is to use lots of large fans set just under your hearing range to get air in and out as efficiently as possible.

What percentage does Catalyst Control Center say the fans are being powered at? I just ordered an Arctic Cooling Acellero Xtreme for my 6870, (3 92mm fans) CCC says the temperature is hovering at 39c with the fans at 23% in my test bench. They become audible with your ear an inch away in the early 30's. Inside a case you can probably get away with 40s or 50s.

I haven't checked during games but a few Firefox windows, Windows Live Messenger and Steam never spike over 10%.


A: Low 40s
B: I just installed it yesterday, I'm still toying with the settings to find the best balance. (Both fans are using the same header.)
C: I configured my motherboard to max out the speed to 50% at 60c. This morning I turned the maximum speed down to 40%. The temperature increased into the 50s.

Gotcha, my logic was that heat from my video card would possibly heat up my CPU or CPU cooler, but I guess it doesn't work like that.

Mine says 41° at idle at 24% fan speed, it can however go as high as 55-60° doing video and maybe flash. Haven't been checking the fan speed as I generally use another hardware monitor that doesn't show the GPU fan speed.

Would be interested in your impressions after you've played with it a bit more.

If it's dumping its waste heat into the case without directed airflow then that warm air could be reaching that cooler and reducing its efficiency.

Sounds like we've found room temperature. Now we need a way to see the temps and speeds while we game.

I'd love to share them.

Personally I am very comfortable as long as the temps are under 50C. When it gets past 50C - I don't like to see that... though I think many of the SPCRers have no qualms taking it up to 60C or even further in pursuit of silence.

Tweaked and rebooted repeatedly this morning until I couldn't take the boredom. Absolute silence ended ~675 RPM per fan with my naked ear an inch away. Best cooling to noise results came when I configured the fans to run at the same percentage as the celcius temperature value of the CPU. (Range set from 40c to 60c) At the time of this posting the CPU fans averaging 725 RPM, CPU load is averaging 7%, processor speed is averaging 850 MHz, and temperature is staying at 38c. (With the CPU locked at 2.8 GHz the temps were around 43c with the same load)

Please note that I'm using a test bench, it doesn't trap and recycle warmed air like a tower or desktop case!
And as I said, the Mugen 2 fans are silent up to 675 RPM, so I can barely hear them right now. In a walled case they would probably be sufficiently muffled.

(Update, February 11th: Today I set the fans to stay close to 675 RPM, the idle temperature is now 37c)

System specs:
CPU - AMD Phenom II x6 2.8 Ghz
Motherboard - Asus Crosshair IV Extreme
Case - Lian Li T60

Are there any measurements of the associated motherboard heat that comes with using a tower heatsink (e.g. Mugan 2) instead of a top-down blower (NH-D14)?

In my experience, down-blowers don't cool the CPU as well as towers, but the surrounding components run MUCH cooler.

The best example I can think of for this is Scythe's Rasetsu, which is the down blowing version of its Yasya tower. The Rasetsu gets runs a few degrees hotter and the fan runs louder at full load, but the Yasya doesn't have any effect below the fins.

Strangely the Noctua NH-C14 seems to be an exception to this rule... though probably there is an explanation for it that will become evident in some new Noctua Tower equivalent that comes out and cools even better.

Though I swear... with the direction that chips are taking... it seems like the future may hold less and less need for super coolers.

In dual fan mode that's a given, but there's no upright version of it with the same fin stack dimensions.

This is probably an overstatement. Down-blowers tend to be better but it depends partly on the details of the tower design. Some have deflectors to direct airflow to the board, and others position the fan(s) low enough that peripheral airflow reaches the board. Also, the fan speed matters a lot. If the fan is running as slowly as we often run them for min noise (say 5~600rpm or lower), even if blowing down, the amount of cooling provided is very small.

It would be great if you guys included motherboard temperature measurements in the SPCR tests

Fans can be upgraded and vibration-dampened to reduce noise at higher output, but sometimes the board design prevents any downward airflow from having an effect, like a northbridge cooler that had a nameplate on top which acted as a roof.