Archive Home arrow Guides arrow Best CPU Cooler Performance: Intel Q3-2010
Best CPU Cooler Performance: Intel Q3-2010 E-mail
Reviews - Featured Reviews: Cooling
Written by Olin Coles   
Wednesday, 17 November 2010
Table of Contents: Page Index
Best CPU Cooler Performance: Intel Q3-2010
CoolIt Vantage ALC VAN-R120
Deep Cool Gamer Storm Heatsink
Prolimatech Megahalems Heatsink
Scythe Mugen 2 SCMG-2100
TITAN EVO TTC-NK85TZ/CS2 Cooler
Thermalright Venomous-X RT Cooler
Xigmatek Aegir SD128264
Zalman CNPS10X-Performa Cooler
CPU-Cooler Preparations
Heatsink Test Methodology
Heatsink Performance: Stock Cooling Fan
Heatsink Performance: Dual Noctua Cooling Fans
Heatsink Performance: High-Output Fan
CPU Cooler Final Thoughts
Best CPU Cooler Conclusion

Best CPU Cooler Conclusion

IMPORTANT: Although the rating and final score mentioned in this conclusion are made to be as objective as possible, please be advised that every author perceives these factors differently at various points in time. While we each do our best to ensure that all aspects of the product are considered, there are often times unforeseen market conditions and manufacturer changes which occur after publication that could render our rating obsolete. Please do not base any purchase solely on our conclusion, as it represents our product rating specifically for the product tested which may differ from future versions. Benchmark Reviews begins our conclusion with a short summary for each of the areas that we rate.

This article offered a few surprises for me, yet none of them came from the new products we tested. Previous articles in this series used an early-production 2.66GHz Intel Core i7-920 processor, which was replaced with a new 2.8GHz Core i7-930 CPU for this project. Using the same testing methodology with higher speeds and temperatures, we witnessed a small change in our overall results. These weren't major changes mind you, especially since the same three cooler dominate the top rankings, but it was the order that these products finished that had changed. In previous tests with the i7-920 at 3.8GHz, the Megahalems consistently out-cooled the Venomous-X while the Mugen 2 trailed them both. This time we used an i7-930 at 4.0GHz, and the Venomous-X beats the Mugen 2 which is trailed by Megahalems. It's madness!

I'm actually disappointed by the limited number of new products received for testing in this article, and particularly frustrated with Thermalright for shipping us the Silver Arrow cooler three months after it was repeatedly requested only to arrive the day before an already long-delayed publication of this article. No, this quarters' Best CPU Cooler Performance project wasn't nearly as large and exciting as previous editions in our series, and the upcoming uncertainty surrounding Intel's Sandy Bridge socket didn't help. Fortunately the upcoming Intel Sandy Bridge LGA1155 (socket H2) on P67/H67 motherboards will use the same mounting holes as LGA1156 (socket H) for P55/H55/H57/Q57 motherboards, so no new mounting clip systems will need to be designed.

I've also learned that processor architecture can have a huge impact on heatsink performance. I'm not referring to speed or voltages here, because those factors are a given when it comes to cooling. What I'm referring to is how the 45nm Intel Bloomfield Core-i7 is going to have a 'heat signature' area that differs slightly from 32nm Gulftown. In fact, Gulftown's 248mm2 die package is closer to a Lynnfield LGA1156 Core-i7 CPU. Those heatsinks with a larger contact surface (and heatpipe base) will best serve 45nm AMD Phenom-II processors with a 258mm2 die or 45nm Intel Core-i7 quad-core 263mm2 Bloomfield CPU's. Essentially, it's important to research the cooler's physical information in addition to performance results when you're shopping for a CPU cooler. It's not a one-size-fits-all heatsink market, and the biggest cooler doesn't always provide the best performance. With these consideration in mind, I will offer several different product suggestions based on these test results and my past experience.

High-Performance Overclocker Heatsinks

These heatsinks are suggested based on quality, value, and cooling potential:

  1. Thermalright Venomous-X RT: The new 'Pressure Vault' mounting system offers incredible contact pressure, and pairs a polished contact surface to densely packed heatsink that supports two 120mm cooling fans. The Venomous-X RT kit costs $68, and includes universal Intel and AMD AM2/AM2+/AM3 mounting hardware along with a 120mm cooling fan. As a standalone CPU cooler Venomous-X delivers on its years of past design experience, and placed at the top of every single test we performed on it.
  2. Scythe Mugen-2 SCMG-2100: Just barely trailing behind is the $40 Mugen 2 CPU cooler, which already includes AMD mounting hardware and a premium Scythe Kaze-Jyuni Slip-Stream fan... all for only half the cost of most high-end options. The Mugen 2 heatsink has already received our Editor's Choice Award honorable mention pertaining to value, and it consistently finishes at the very top. The Scythe Mugen-2 Revision B model SCMG-2100 heatsink is identical to the first version, model SCMG-2000, but SCMG-2100 includes a newly developed F.M.S.B. (Flip Mount Super Back-Plate) for more convenient mounting onto the motherboard and the thermal paste has been replaced by SCYTE-1000.
  3. ProlimaTech Megahalems: Using the best mounting system we've ever tested this heatsink delivers extraordinary contact surface pressure between the processor and polished base. The results produce the best cooling performance we've seen, even beating their new Super Mega heatsink. As of November 2010 the ProlimaTech Megahalems was available for $62 at FrozenCPU and NewEgg. AMD users will also want the AM2/AM2+/AM3 mounting kit from FrozenCPU for an extra $10. Of course, adding two high-output cooling fans will make Megahalems unstoppable.
  4. Zalman CNPS10X-Performa: Finishing second in stock results, third with dual silent fans, and fifth when a high-output fan is attached, this inexpensive heatsink really delivers excellent bang for the buck. Priced for only $40 at NewEgg, this could be the poor man's answer to the more expensive Megahalems and Venomous-X while being half as large as Scythe's Mugen 2. The Zalman CNPS10X-Performa fits nearly all Intel and AMD sockets, and comes with a pouch of Zalman Super Thermal Grease ZM-STG2M.
  5. Cogage TRUE-Spirit: Priced at $40, this heatsink takes the Thermalright Ultra-120 Extreme (TRUE) design essence and fits it into a smaller more affordable CPU cooler. Our tests of the Cogage TRUE Spirit has always delivered impressive performance, and although only Intel LGA1156/1366 sockets are supported, an optional AMD AM2/AM2+/AM3 mounting system is available for $10.

Enthusiast Heatsink Considerations

If you're not planning to increase vCore voltage to your processor, here are some additional recommendations that could fit your build:

  • Xigmatek Balder SD1283: Since the Xigmatek Aegir SD128264 and Titan EVO appear to be impossible to find for sale online, I've selected the Xigmatek Balder SD1283 as a great alternative. Based on the original HDT-S1283 design, Balder offers some subtle improvements that help it maintain top-rated cooling performance in our tests. While the Balder kit includes a bolt-through mounting kit and costs $40, the original Xigmatek HDT-S1283 model uses Intel push-pin clips and still sells for $35.
  • GELID Tranquillo CC-TRANQ-01-A: GELID has declined to send us a sample for our round-up comparisons, explaining that their product is designed for low-noise application and not top-end overclockers. We disagree. Paired with a high-output cooling fan, the Tranquillo heatsink could offer an affordable CPU cooler for only $40.
  • Zaward Vapor-120 ZCJ013: A sturdy bolt-through mounting system combined with dimpled heatsink fins and 'Golf' fan blades keep performance up and noise down. Selling for $45 this could offer ample cooling protection to moderately overclocked processors, plus it's compatible with every motherboard socket currently in existence.
  • Thermaltake V1 CL-P0401: Back in stock with a reduced price of $52 and now compatible with all current Intel and AMD processors, I'm re-suggesting the Thermaltake V1 after originally testing it to perform at unbelievably cool levels. Sure, it's not your average tower-style heatsink, and it's nearly 150mm tall, this copper calamity reduces temperatures as well as the others in this section. The only difference is that the V1 looks a lot more unique, and sometimes that makes all the difference.
  • Scythe Ninja-3 SCNJ-3000: Although the results of our Ninja-3 tests have yet to be published, this heatsink embodies the test-proven design of the Mugen-2 and other heatsinks. Scythe products have already earned many awards from us, and for $52 this not-so-subtle CPU cooler could be the next in line.

A Word About ALC Water Coolers

In this article we tested the CoolIt Vantage ALC, a liquid cooling product available at NewEgg for $106. Much like the CoolIt Domino and ECO coolers we've previously tested, the CoolIt Vantage ALC performed considerably well with nearly no audible noise under full load. Beyond cooling performance, they also added several features not found on other All-In-One Liquid Coolers such as temperature alarms, real-time status information, and color LCD screens. What I find surprising is that most consumers are still fully unaware of the CoolIt product line, and instead opt for re-packaged Asetek clones.

I've recommended a lot of air cooling products over the past few years, but I think it's finally time to suggest a few liquid cooling solutions. I like the Corsair Hydro H50 we previously tested, but for $80 this re-branded cooling solution offers none of the monitoring features or functionality that a $65 Domino ALC system does. Although the ECO R-120 is essentially the same as a Hydro H50, but for $72 you get the same performance for less money. It's so frustrating to see consumers purchase their products based on name recognition instead of performance. This appears to be the case with Corsair's Hydro H70 ALC for $105, which is just an H50 with larger radiator and lacks any of the monitoring and feedback display you get with the $106 Vantage. This isn't a company endorsement, it's just common sense.

So what do you think we should test in the next CPU cooler article? Leave comments below, or ask questions in our Forum.

There are so many cooling products on the market that deserve our attention, so don't think for a minute we've intentionally left someone out. In order for this series to hit the target audience with the most relevant product coverage, Benchmark Reviews needs your feedback. Let us know what you want to see tested, and tell us where to find it!


Related Articles:


 

Comments 

 
# RE: Best CPU Cooler Performance: Intel Q3-2010mihai 2010-11-17 05:12
good review as usual, a+ from me.
Guys where is V6GT? from precedent review was a very good performer,
Report Comment
 
 
# RE: RE: Best CPU Cooler Performance: Intel Q3-2010Olin Coles 2010-11-17 08:06
One of our other writers has the V6GT, which is why I didn't test it in this comparison (although I probably should have asked for it, since so many other manufacturers failed to ship samples). The Cooler Master V6GT review offers a pretty good comparison to the top heatsinks, and should give you an idea where it places.
Report Comment
 
 
# Excellent!RealNeil 2010-11-17 06:36
Yeah,......
This is a good one guys. Something that I can link to and refer people to that have questions about CPU cooling solutions. While there is always gonna be someone who brings up questions about why wasn't my 'Blah-Blah Cooler' included, you have a good assortment here and tested them all fairly.

Thank You
Report Comment
 
 
# RE: Excellent!mihai 2010-11-17 15:18
calm down...blah blah .....
in this review was used different hardware...so the results could be different.....anyway ...
Report Comment
 
 
# RE: Best CPU Cooler Performance: Intel Q3-2010Holy_COW 2010-11-17 10:06
No Corsair A70?

Other reviews show it as a good HSF, but will be great to see it included on a comprehensive comparison as this one

:(
Report Comment
 
 
# CoolIt is better than Corsair HydroOlin Coles 2010-11-17 10:46
Obviously you didn't read the last two paragraphs in the conclusion.
Report Comment
 
 
# Not a Corsair HydroCPT. Obvious 2010-11-17 16:46
Obviously you didn't read his post because he asked about the A70 which is an air cooler.
Report Comment
 
 
# RE: Not a Corsair HydroOlin Coles 2010-11-17 16:59
You're right! But if that's the case, the A70 is a direct clone of countless outher HDT coolers. Take your pick of HDT coolers with four heatpipes that can fit two fans. We didn't receive this item for review, which is why it wasn't included.
Report Comment
 
 
# Corsair A70Phil 2010-11-19 19:35
I would also like to see it paired against other intel heatsinks, but I would also like to see one in the Amd review as the Corsair A70 is only $48 now and currently on newegg for $33 with rebate. I haven't seen any good reviews with an amd processor.
Report Comment
 
 
# RE: Best CPU Cooler Performance: Intel Q3-2010PSM 2010-11-17 16:39
Congrats on this fine comparison review. This would be enough for decision making for new buyers, at least till SandyBridge based chips come out in 1stQ of next year. But Thermaltake Frio, which is one worthy cooler, should have been included. It also comes with possibly the best pair of default fans. Keep up the works.
Report Comment
 
 
# RE: Best CPU Cooler Performance: Intel Q3-2010Athlonite 2010-11-17 21:24
hows about group test of a range of prebuilt water cooling solutions
Report Comment
 
 
# RE: RE: Best CPU Cooler Performance: Intel Q3-2010Olin Coles 2010-11-17 21:31
We're already working on that, and should publish about a week after Thanksgiving.
Report Comment
 
 
# Noise readings?Shane 2010-11-18 16:46
First off, thanks for the awesome guide. It's this and it's predecessors that got you guys in to my Bookmark Bar.

Are there any plans to record noise levels? Obviously (based on your review) using two Noctua fans is a silent or near silent solution, and strapping on that Yate Loon is a recipe for distracting the neighbors, but what about the stock fans? I'm willing to spend money on a decent solution (namely the SCMG-2100) but may not have the budget for two super fancy fans at ~$15 a pop. Plus, if spending $30 is only going to get me 2db less than the included Slipstream, it's just not worth it.
Report Comment
 
 
# RE: Noise readings?Olin Coles 2010-11-18 17:15
Thank you for your feedback! The 'Special Purpose' version of the Yate Loon fan that I use could break down kidney stones with its ear-piercing noise. In comparison to the Slipstream fan that comes with a Mugen 2, they're night and day. The Scythe Slipstream fans are slightly more audible than the Noctua fans, but not by much. As such, they don't push nearly as much air as the Yate Loon, and are much closer to the Noctua fans.
Report Comment
 
 
# Great articleCraig 2010-12-05 23:22
Thanks for a great article. I appreciated the time you took explaining how thermal paste works in relation to the transfer of heat from the cpu to the cooler. I have to sheepishly admit to not thinking much about after market coolers, as I do not overclock my systems, but after reading this article I may have to rethink things. If you could, can you tell me in general how much of an improvement would I realize in using one of the top three coolers vs. stock Intel or AMD solutions?
Have Fun!
Report Comment
 
 
# RE: Great articleOlin Coles 2010-12-06 11:33
Hello Craig:

It's been a while since we tested an overclocked i7 CPU with the stock cooling solution... generally because it's too dangerous. Here's an example from the last time we did: benchmarkreviews.com/index.php?option=com_content&task=view&id=371&Itemid=62&limit=1&limitstart=11
Report Comment
 
 
# RE: RE: Great articleCraig 2010-12-06 15:22
Thanks for responding so quickly Olin. I read the article you recommended and it looks as if I would see a difference of 15 - 20 degrees vs. a stock heat sink and fan, pretty significant.
Report Comment
 
 
# RE: RE: RE: Great articleOlin Coles 2010-12-06 15:36
Yes, it's a very significant difference in cooling performance. Also consider that those were at the stock voltage and clocks, and the difference would be considerably more on an overclocked system.
Report Comment
 
 
# My Scythe Mugen 2 results versus yoursJohan Niemi 2010-12-22 12:25
Hey Olin, how do you think these results compares to yours?

Scythe Mugen 2, default fan at maximum speed
i7 950, 4GHZ, 1.4Vcore
Artic Silver 5
Everest 5.5
Ambient room temp: 23c

Stress CPU and FPU selected and run for 30 min. All temps are from the statistics panel, from the average column and with an ambient room temp of 23.

CPU: 57,6
Core1: 79,9
Core2: 77,3
Core3: 76,4
Core4: 73,1

I don't know how you get your temp 41.39, but I assume you add all core temps, divide by four and subtract ambient room temp? In my case that's 53,7c. That's 12.3c more than your reading. Can that simply be explained by different motherboards?
Report Comment
 
 
# RE: My Scythe Mugen 2 results versus yoursOlin Coles 2010-12-22 13:05
Hello Johan:

You've got the right idea: Core1-4 are averaged and then subtracted from the ambient temperature. I don't take the maximum Core readings, however, I allow Everest/Aida64 to average them for the final ten minutes of the test run.

Motherboard sensors and BIOS calibration are different between all motherboards; even those of the same model. Another big difference is voltage regulation, where some motherboards are analog while others (such as MSI) are digital and more precise. This makes it difficult to match results between systems. What's important is scaling between heatsinks tested on the same system.
Report Comment
 
 
# RE: RE: My Scythe Mugen 2 results versus yoursJohan Niemi 2010-12-22 13:37
Thanks. I should correct my above post, the Vcore was set to 1.34375 which makes my temps seem even higher.

Scaling between heatsinks then, this might be interesting.

Prime95 blending for 1h (1.33125Vcore, 4GHz) causes a max temp of 94c with my Xigmatek HDT-S1283. With the Scythe Mugen 2 it's 88c. The readings are the max temp for the hottest core. The max temp of the coolest core was 84 with my S1283 and 81 with the Scythe Mugen 2.

6c gained on the hottest core. 3c gained on the coolest core.

That's not really a whole lot gained when you take into account that HDT coolers should be really bad with i7, and Scythe Mugen 2 is the runner up in "Best CPU Cooler Performance: Intel Q3-2010".

Am I expecting too much?
Report Comment
 
 
# RE: RE: RE: My Scythe Mugen 2 results versus yoursOlin Coles 2010-12-22 13:47
If you're trying to compare against my results, you'll need to use the same software and methodology. I use Everest to load all cores to 100%, and use the built-in averaging system to report the temperatures.

Installation and TIM application play a big role, and if you're using too much paste then your temperatures would be higher than mine. You also need to take into account the heatsink orientation (vertical/horizontal), which makes a considerable difference.
Report Comment
 
 
# RE: RE: RE: RE: My Scythe Mugen 2 results versus yoursJohan Niemi 2010-12-22 14:07
I was trying to compare against you're results in the first post. What exactly do you mean by letting Everest "average them for the final ten minutes of the test run"? In my first post I used Everest's stress test, ran it for 30 min, then I took the the "Average" numbers and manually added them, divided by four and subtracted the ambient temp. Is there a hidden function somewhere in Everest that averages the average temps?

In my second post however I wasn't trying to compare against your results but my own. I used Prime blending to generate the heat, and then noticed the max temps with Real Temp.

The Xigmatek S1283 generated these max temps: 94, 92, 90, 84

The Scythe Mugen 2 generated these: 88, 86, 84, 81

Is that what your would expect?
Report Comment
 
 
# RE: RE: RE: RE: RE: My Scythe Mugen 2 results versus yoursOlin Coles 2010-12-22 14:19
Yes- that appears approximately correct. Please keep in mind that heatsink orientation still plays a big role in performance. See notes in the 'CPU-Cooler Preparations' section.
Report Comment
 
 
# RE: RE: RE: RE: RE: RE: My Scythe Mugen 2 results versus yoursJohan Niemi 2010-12-22 14:49
Alright, I was hoping for a bigger improvement since I'm getting very close to TJmax (100c). I have only mounted the Mugen 2 with the fan pushing (not pulling) air towards the back of chassis.

Correct me if I'm wrong, but the fact that my Rampage III Extreme reports temps so much higher than your motherboard, wouldn't that hold my OC back? I mean, you can push your Vcore so much more before reaching TJmax than I, because your temps are reported lower.

Right now with a Vcore at 1.35625 my max core temp is 91, only 9 degrees from max. I'm doubting that I'll be able to reach 1.4v. Now if my reported temps were some 12c lower, that would leave me with some headroom.
Report Comment
 
 
# RE: RE: RE: RE: RE: RE: RE: My Scythe Mugen 2 results versus yoursOlin Coles 2010-12-22 14:53
I cannot estimate what your OC could be on that motherboard or the safe limits, although temperatures could be further reduced. See notes in the 'CPU-Cooler Preparations' section for the details.
Report Comment
 
 
# RE: RE: RE: RE: RE: RE: RE: RE: My Scythe Mugen 2 results versus yoursJohan Niemi 2010-12-22 15:22
I know, perfect pressure and perfectly applied TIM, etc. can reduce temps a bit. I have read that section several times during the last years. :)

But changing motherboard can too.

So my question is: Shouldn't lower reported temperatures result in possible higher OC's? If yes, and you can have lower reported temps by changing motherboard, then it should be possible to reach higher OC's just by changing motherboard.

So manufactures could use the fact that a given product is known for low readings as a sales argument. Seems a little silly.
Report Comment
 
 
# RE: RE: RE: RE: RE: RE: RE: RE: RE: My Scythe Mugen 2 results versus yoursOlin Coles 2010-12-22 15:51
I'm specifically referring to the 'Heatpipe Directional Orientation' section that's rather new. Turning the heatsink 90-degrees dropped my temps by almost 4C on some coolers.
Report Comment
 
 
# RE: Best CPU Cooler Performance: Intel Q3-2010Johan Niemi 2010-12-22 16:00
That is something! I'll give it a try and blow air up towards the top of chassis. If only the screwing didn't have to be done from the back of the board, it's rather cumbersome.
Report Comment
 
 
# RE: RE: Best CPU Cooler Performance: Intel Q3-2010Olin Coles 2010-12-22 16:10
You should try doing it three times apiece per heatsink.

At any rate, make sure that the heatpipe rods travel as horizontal as possible. You might already have them that way... on the Scythe Mugen 2 the heatpipes should be on the left and right, as should the fans. On HDT coolers it's usually opposite.
Report Comment
 
 
# RE: Best CPU Cooler Performance: Intel Q3-2010Dietmar Habeck 2011-06-14 21:31
Your test results ever appreciated
Given your findings that different CPU's have different heat signatures
I am not finding any comparative test past the last Q3 2010 and it is June 2011.Did you stop them?
I am looking for a good cooler for the Intel LGA 1055 platform i5 2500K CPU
thnx
Report Comment
 
 
# RE: RE: Best CPU Cooler Performance: Intel Q3-2010Olin Coles 2011-06-14 21:41
Hello Dietmar:

While Benchmark Reviews has published several CPU cooler articles, we have not produced another comparitive roundup like the 'Best CPU Cooler Performance' series. This is for two reasons: 1) the latest Intel CPUs (Sandy Bridge) run significantly cooler, and do not require the same high-end heatsink as before, and 2) there haven't been that many new heatsinks introduced.
Report Comment
 
 
# Ven X pressurePapant 2011-07-26 22:59
How much pressure did you apply on the ven x? Did you screw it tightly?
Report Comment
 
 
# RE: Ven X pressureOlin Coles 2011-07-27 07:43
The 'Pressure Vault' was secured as tightly as it could be.
Report Comment
 
 
# RE: RE: Ven X pressurePapant 2011-07-27 07:46
did it wiggle a lot? Mine wiggles a lot and there has been a lot of buzz about socket damage caused by applying max pressure.
Report Comment
 
 
# RE: RE: RE: Ven X pressureOlin Coles 2011-07-27 07:48
No, this unit was very secure and locked into place. It is possible to cause socket damage if you're not careful, but we did not experience this.
Report Comment
 

Comments have been disabled by the administrator.

Search Benchmark Reviews
QNAP Network Storage Servers

Follow Benchmark Reviews on FacebookReceive Tweets from Benchmark Reviews on Twitter