1366 CPU Heatsinks Round 1

1366 CPU Heatsinks Round 1
Thermalright Venomous X – Noctua NH-U12P SE2
Prolimatech Mega Shadow – Prolimatech Megahalems


Introduction

Welcome to round one of my air-cooling heatsink round ups. This round up features some of the big players in the heatsink field. With tried and tested heatsinks up against Thermalright’s latest offerings the Venomous X. Each of the manufacturers has gained respect from air-cooling enthusiasts for a whole range of reason. But today we’re putting they differences aside while we test each of the four heatsinks to find out who owns the crown.


http://farm5.static.flickr.com/4063/...2518ddc7be.jpg

http://farm5.static.flickr.com/4023/...cc7d9fa4_o.png


Before we crack on, I would like to thank Prolimatech for sending me review samples of two of their coolers (Mega shadow & Megahalmes) and thanks to Noctua for providing me with a discounted NH-U12P SE2.


Thermalright Venomous X

Since 2002 Thermalright has been set at the high end of the heatsink market with both design and functionality. And since then in recent time Thermalright has been the undisputed kings of air-cooling since the introduction of the Ultra-120 Extreme back in 2007.

An Overview of what is included:

•Venomous X Heatsink
•Instructions
•Thermal Interface Material (TIM for short)
•Mounting Hardware
•2 Sets Of Fan Clips
•4 Fan Anti Vibration Stripes
•Thermalright Sticker

http://farm5.static.flickr.com/4012/...82cd238bf3.jpg

http://farm5.static.flickr.com/4070/...0b85933165.jpg


The first thing you notice about the Venomous X is the box. Unlike most other Thermalright products, which arrive in a plain brown cardboard box, Thermalright have put some time, money and effort into the box. Which is fantastic for the first impressions of the cooler.


http://farm5.static.flickr.com/4035/...63e71fe957.jpg

http://farm5.static.flickr.com/4015/...30aa804e58.jpg

http://farm5.static.flickr.com/4023/...f8a546b580.jpg

http://farm3.static.flickr.com/2695/...0f85136f0a.jpg


Once you get inside of the box, you will be greeted with the accessory pack, the accessory pack contains all the mounting hardware you will need to mount the Venomous X. The accessory pack comes in glossy black box, which is rectangle is shape.

Once you’ve removed the accessory pack from the box, you will find the instructions and the free included Thermalright sticker.

The heatsink is the last remaining piece in the box, wrapped in clear plastic bag then imbedded firmly on a foam casing. The nickel-plated heatsink weighs in at 755 grams and its dimensions without fans are 127 x 63 x 160mm (W x D x H). Six heat pipes run though the 47 fins, which are 0.60mm thick and spaced out at 1.5mm spacing’s.


Prolimatech Mega Shadow

Prolimatech founded in 2008 by a team of experienced thermal engineers. Based in Taiwan with a quest to satisfy every overclocker's and enthusiast's needs. Prolimatech stands for Professionalism exceeding beyond all Limits.

An Overview of what is included:

•Mega Shadow Heatsink
•Instructions
•Thermal Interface Material (TIM for short)
•Mounting Hardware
•2 Sets Of Fan Clips
•Intel Backplate

http://farm3.static.flickr.com/2796/...19af3f66bf.jpg

http://farm3.static.flickr.com/2483/...223a73d9cd.jpg


The Mega Shadow gets delivered in an unusual style of box, with a removable top that has clear plastic window cut out. The box has a clean style with a basic white background and a metallic mega shadow drawing on the front and back.


http://farm3.static.flickr.com/2747/...290a09e754.jpg

http://farm5.static.flickr.com/4025/...1b3e7f645f.jpg

http://farm5.static.flickr.com/4004/...9a000a0101.jpg

http://farm3.static.flickr.com/2786/...f0f357b566.jpg


Once you lift the top of the box off, you will find the accessory pack to one side of the heatsink. The accessory pack contains all the mounting hardware you will need to mount the Mega Shadow. The accessory pack comes in glossy white box, which is rectangle is shape.

The heatsink is the last remaining piece in the box, wrapped in clear plastic bag then imbedded firmly on a foam casing. The black nickel-plated heatsink weighs in at 790 grams and its dimensions without fans are 130 x 74 x 158.7mm (W x D x H). Six heat pipes run though the 45 fins, which are 0.75mm thick and spaced out at 1.5mm spacing’s.


Prolimatech Megahalems

Prolimatech founded in 2008 by a team of experienced thermal engineers. Based in Taiwan with a quest to satisfy every overclocker's and enthusiast's needs. Prolimatech stands for Professionalism exceeding beyond all Limits.

An Overview of what is included:

•Megahalems Heatsink
•Instructions
•Thermal Interface Material (TIM for short)
•Mounting Hardware
•2 Sets Of Fan Clips
•Intel Backplate

http://farm3.static.flickr.com/2739/...636674036f.jpg

http://farm3.static.flickr.com/2483/...223a73d9cd.jpg


The Mega Shadow gets delivered in a regular rectangle box, with the top being swung open. The box has a clean style with a two-tone carbon fiber effect graphic, ¾ dark blue and ¼ light blue. A white wire frame drawing of the Megahalems sits nicely on the front of the box.


http://farm3.static.flickr.com/2720/...46c50c9d44.jpg

http://farm3.static.flickr.com/2704/...8ac287f479.jpg

http://farm3.static.flickr.com/2677/...542c7da1ab.jpg

http://farm5.static.flickr.com/4052/...9de4460457.jpg


Once you lift the top of the box off, you will find the users manual and Intel Backplate. The accessory pack and the heatsink are stored in a removable brown cardboard shell. Once you remove the removable inner cardboard housing the accessory pack can be found in the fold out pocket on the bottom of the heatsink. The accessory pack contains all the mounting hardware you will need to mount the Megahalems.

The heatsink is the last remaining piece in the box, wrapped in bubble wrap. The nickel-plated heatsink with a scratch-resistant Stainless Steel top weighs in at 790 grams and its dimensions without fans are 130 x 74 x 158.7mm (W x D x H). Six heat pipes run though the 45 fins, which are 0.75mm thick and spaced out at 1.5mm spacing’s.


Noctua NH-U12P SE2

Noctua is a joint Cooperation between Austrian computer destruction company Rascom and Taiwanese Cooling specialists Kolink. With both companies Specializing in different parts of the industry, they knowledge is combined bringing premium heatsink with excellent customer service. “Noctua”, Meaning the Little Owl, the Greek goddess of Athens. She Represents Wisdom, Strategy and Science.

An Overview of what is included:

•4 -Duel Heat Pipe UD12P heatsink
•2 Noctua NF-P12 Fans
•SecuFirm Multi Socket Mounting System
•NT-H1 Thermal Compound
•Philips Screwdriver
•2 Sets Of Fan Clips
•AMD Mounting Kit
•Intel Mounting kit (775, 1156, 1336)
•2 Low Noise Adapters (Fan Extension Cable Which Lowers the Voltage)
•2 Ultra Noise Adapters (Fan Extension Cable Which Lowers the Voltage)
•Installation Manual For AMD & Intel

http://farm5.static.flickr.com/4012/...d21215bf6f.jpg

http://farm5.static.flickr.com/4071/...8b05a7f87e.jpg


The Heatsink along with the fans and accessories gets delivered in the sleek square cardboard box. The outside of the box has all the usual features, Reviews, Information and diagrams. But there is one major feature, which really stands out, and this is the graphics and specification on the top panel section of the box.

The specification is for both the heatsink and the fans; bearing Power Consumption, Airflow and Noise against other things. The 3D White Line drawing with the maroon colored background brings a sense of well-engineered produce.


http://farm5.static.flickr.com/4071/...35e8a46c3c.jpg

http://farm5.static.flickr.com/4009/...be816d25bc.jpg

http://farm3.static.flickr.com/2435/...207c976ceb.jpg

http://farm5.static.flickr.com/4058/...c1116b8c41.jpg


The accessories and the remaining parts of the cooler are packed separately, In addition to the heatsink you will find, Thermal Paste, Mounting Hardware, Installation Manuals and the Seconded of the NF-P12 Fans.
The heatsink is the last remaining piece in the box surrounded by a white cardboard outer shell. The nickel-plated heatsink weighs in at 600 grams and with fans respectively 940 grams. The heatsinks dimensions without fans are 126 x 71 x 158mm (W x D x H). Three heat pipes run though the 36 fins, which are 0.60mm thick and spaced out at 2.5mm spacing’s.


Thermal Test Specification

A 5-mount test with 1hour with each mount, 40 minutes ‘warm up’ and 20 minutes of temperature logging.

•The 1366 Test Setup is as follows: Intel Core i7 920 D0 which is at stock 2.66GHZ (no lapping) running on a Gigabyte UD5 Extreme Motherboard. 6GB (3x2GB) of Patriot Viper DDR3 1600MHz Memory - the motherboard cooling is stock with no extra airflow. Video card is a BFG 9800GXS, also on stock cooling but the outer casing removed. The power comes from a Zalman 850W Modular PSU; Samsung 500GB hard drive split into 2x250 GB partitions. Test case is a Le Chuck Test Bench, which has horizontal mounting.

•Each cooler is tested with one, two and no fans. Each is repeated 5 times with 40 minutes ‘warm up’ and 20 minutes of temperature logging.

oEach cooler is run with no fans (Passive), this is to see its thermal threshold. It’s ability to hold and disperse heat without airflow.
oOne-fan testing is deployed into 2 sections. The fan is run at 1300 and 800rpm controlled via a Crystalfontz 635.
oTwo-fan testing is deployed into 2 sections. The fans are ran at 1300 and 800rpm controlled via a Crystalfontz 635. Where applicable the fans are lined into a push pull configuration. This means one fan is pushing air though the heatsink and the other is pulling it thought the other side. Adding more pressure and hopefully gaining better temps.

•Each cooler is tested at overclocked settings with one, two and no fans. Each is repeated 5 times with 40 minutes ‘warm up’ and 20 minutes of temperature logging. The overclocked CPU settings are 20x200MHz which is equal to a modest 4GHz overclock.

oEach cooler is run with no fans (Passive), this is to see its thermal threshold. It’s ability to hold and disperse heat without airflow.
oOne-fan testing ran at 1300rpm controlled via a Crystalfontz 635.
oTwo-fan testing ran at 1300rpm controlled via a Crystalfontz 635. Where applicable the fans are lined into a push pull configuration. This means one fan is pushing air though the heatsink and the other is pulling it thought the other side. Adding more pressure and hopefully gaining better temps.

•The same fans which are used on each heatsink are Noctua's NF-P12 120mm Fan.

•The thermal interface material (TIM) I use is MX-2 like most other enthusiasts and testers alike, with it working as soon as you apply it (no break in time like AS5). Also being non conductive, conductive could become a problem with multiple applications per test and multiple coolers.

The TIM application method I use is the dot in the center. (Small pea sized dot applied to the center of the CPU, then when block is Compressed TIM is spread evenly across the cores of the CPU), Recorded data is only used if the TIM application is good after inspection on remount (my way of validating my results). The CPU is cleaned after every mount using Arctic Clean.

•I have 18 temperature probes across the test bench, 10 Dallas DS18B20 Digital one-wire sensors positioned as follows; 2 x Heatsink Fins, 4 x Air In, 4 x Air Out and 4 Intel DTS in the CPU.

•Temperature logging is recorded via my Crystalfontz 635, Using they WinTest V1.0 software for the Dallas one-wire sensors (Air). WinTest automatically logs temps and outputs them to .CSV files. OCCT V3.1.0 is used to log the core temps using the CPU’s built in DTS sensors, which are also automatically outputted to .CSV files. The .CSV are loaded in to Excel where the data in analyzed and interpreted into Graphs.

Thermal Test Results (Stock)

http://farm3.static.flickr.com/2694/...32b0ec45_m.jpg http://farm5.static.flickr.com/4021/...528abf07_m.jpg http://farm3.static.flickr.com/2683/...d4034618_m.jpg http://farm3.static.flickr.com/2737/...6b35a3a1_m.jpg http://farm5.static.flickr.com/4028/...1aa60f0f_m.jpg


Thermal Test Results (4GHz Overclock)

http://farm5.static.flickr.com/4015/...2e421023_m.jpg http://farm5.static.flickr.com/4011/...b8908cdd_m.jpg http://farm5.static.flickr.com/4043/...6e6ce694_m.jpg

Nice one, but you should first apply a huge overclock to the cpu (4ghz at least) and test with different fans, I don't think anybody expends that much money into a top tier HS not to overclock the cpu then...

I agree. You should test at stock, then a mild O/C, and finally max O/C. But its a good start.

I think he has another round or two to follow hence the title so hopefully we will see some overclock temps etc.

grammar is good! haha

What is your native language?

I hope round 1 refers to stock settings for the cpu and not this batch of heatsinks because you realize almost nobody will pay over 50$ for a heatsink to use it at stock settings.

Overclock Temps will be added in the next few weeks, The next round up will feature different heatsink.

^ NIce test dude.

What new heatsink is coming next round?

I would like to see the Tuniq Tower 120 Extreme if you could.

So far i have a Noise Blocker TwinTec, but i'm looking at the Xigmatek Thorshammer, Thermalright Ultra-120 eXtreme Rev.C and A Alpenföhn Nordwand. It all depends how much spare money i have and which manufactures will be willing to support me with samples.

BTW, haven't you thought about only doing this tests for heavily overclocked systems? I mean, doing soooo much work for stock cpu's is kinda pointless IMO, you would get the results you need by just overclocking your cpu and doing the tests.

I would have to agree.

Test them at a high heat load because some of them are made to run with a higher load.

I like the presentation bro, good luck with the tests :)

Quote:

---

Originally Posted by prava

BTW, haven't you thought about only doing this tests for heavily overclocked systems? I mean, doing soooo much work for stock cpu's is kinda pointless IMO, you would get the results you need by just overclocking your cpu and doing the tests.

---

Not everyone Overclocks ;), However you must understand there is over 20 hours of work for each cooler just on stock cooling.

I am working on overclocking temps but the hold up is finding a good median between and light overclock and a extreme overclock.

Maybe you don't understand what xs means. I'll translate it for you: anything but stock. :)

Why would I wanna see some graphs at stock settings that don't do anything for the products involved? Why would I buy one or the other since the difference in minimal? What is the saturation cooling threshold for the products?

Maybe some of these question might help you find the right oc setting.

Very nice presentation, looking forward to seeing the overclocked test. Been looking to upgrade my old TRUE for my i7920 @ 4.0

This roundup is actually really helpful for me - I have the crappy Intel stock fan and even at 2.66GHz my load temps are up to 70C within a few minutes of prime. :(

Quote:

---

Originally Posted by bentleya

Not everyone Overclocks ;), However you must understand there is over 20 hours of work for each cooler just on stock cooling.

I am working on overclocking temps but the hold up is finding a good median between and light overclock and a extreme overclock.

---

Don't get me wrong, mate. The presentation and the work you did in this tests are awesome :up::up: ... but pointless :rolleyes:

Awesome because you put a lot of effort and everything is really well explained and throughly tested...but pointless because if you run any cpu at stock any HS in between the top15-20 will be fine.

So, what is my point exactly? Well, first of all, who needs a bad-ass HS? Somebody who is not happy with his/her temps. Who is, most of the time, that kind of people? Those who overclock the crap out of their cpu. So, why would you invest your time in making such effort in order not to find any real world conclusion? I mean, I have both the NH-U12P, the TRUE, the Ultra-90 and watercooling gear and let me tell you something: I could pick blindly any of those to cool any of my stock cpu's and It would not do any difference at all. If, instead, we overclock the system, I can tell you that there are differences between all those HS because, as we push to the limits, we see who stands over each other once the dust settles.

This would be the same as comparing 5870 vs GTX295 on a 640x480. Do you think it makes any difference which of those pick at that resolution? Do you think anybody who plays at that resolution will pick any of these such expensive cards? This is exactly the same as what you are doing: we are not talking about $20 HS, but about $50-80 ones, so lets make apples to apples and test what really matters: which is overclock to the max.

PS: imagine what would we say to Vapor or Skinee if they tested an i7 920 @ stock clocks... Why they don't even bother about testing stock clocks? Because any cpu-block around the market is more than enough for any stock-clocked cpu you can put it in. Here it's exactly the same.

PSS: just a bit of perspective here, like I said you did an amazing presentation and everything is well written (hell, even for somebody like me who isn't native english speaker it was damn easy and comfortable to go through all the test :up: ). So keep up the good work.

PSSS: you should just pump your cpu to 4.2ghz and test the HS you have around. You would save a lot of time and you would have the most conclusive results...

nice job but if i run on stock i won't be waste my money on them. i should get a 1TB drive to store something.

I would like to see the performance difference between these coolers:

1. IFX-14
2. Noctua D14
3. Megahalems Rev.B
4. True Copper Ultra 120 Rev.C

and you still haven't noticed your grammar and spelling errors. You also copy-pasted the Mega Shadow section to the Megahalems section and forgot to edit the name LOL

Also not showing the units on the graphs is kind of unpro.... why is it the Air / Core Delta? Isn't the graph showing the average temperature? Delta is the difference, not avg.

Venomous X is better than TRUE Cu

Testing Megashadow vs Megahalems is bit useless.. no way that thin nickel plating can change performace. Rest is interesting..

http://translate.google.com/translat...14&sl=ro&tl=en


This should answer some of the questions raised above.

Venomous X and Cogage Arrow review coming soon.

Been waiting for this for a while and they just posted it, so I'd recommend anyone interested check it out: Best CPU Cooler Performance Intel/AMD Q1-2010 @ BenchmarkReviews

can't seem to get google to translate the CONCLUZII - thanks halfwaythere.

Quote:

---

We arrived at the end of this article, important things have been said so far the results speak for themselves. I will of course draw some important conclusions on the twin tower architecture in general and in particular each cooler.

If before I had doubts about the performance or efficiency of this architecture (doubts caused by poor performance of IFX-14 on the 775 platform, but also the lack of performance of this architecture coolers on the market), they all disappeared along with testing of NH D 14. I noticed a very good performance in extreme categories, very difficult to achieve. It is the first time in the history of "modern" when a cooler is the best and at low rpm / fanless and extreme TDP / CFM. A huge contribution to the quality and experience Noctua had but his good behavior IFX-14 Intel Core i7 Platform strengthens the conclusion that the twin tower architecture is among the best solutions in creating a very efficient cooler. Remarks and good behavior in the housing, in terms of a very low airflow, provided that the cooler to be optimized to a certain extent and for low air flow. Using the three fans is another strong point of this architecture, shown in tests done in practice.

I said several times that it is almost impossible to build a cooler to be the best in almost all categories, and those from Noctua almost managed the impossible. Megashadow NH-D14 still win two categories (1 x 1 x 500 rpm and 800 rpm) which represents and weaknesses of this architecture: efficiency is not very good at using a low speed fan / very small. I stated at the beginning of Article main disadvantages of twin tower architecture and we have shown practically in this article. However, the Noctua NH-D14 is closest to the idea of the perfect cooler.

There are plenty to say about the heat generated by the two processors. Note that in the simple case of Q6600's TDP because much less observe an equalization of outcomes. It is clear that the process must be tested for highlighting how coolerele best of their differences, and it is clear that ancient processors no longer find respectable place in a test of coolers. IHS's dimensions, fundamental difference between the existence of two pills in the IHS, to the existence of a single pill, the number of transistors and their density in the case of modern processors, power efficiency floor processor if Intel 45nm, surface pill / pills silica reported in the IHS's total area, the need for a clamping system with high pressure in case of modern processors, different behavior depending on the TDP of the cooler top s are all decisive argument in the correct choice of test platform for coolers. And right now, LGA 1366 platform is the only choice for a cool test results really relevant to the user.

ASUS TRITON 88

Asus has a name "very high" in the field coolerelor but I can say that Triton 88 behaved quite well, especially if we take account of very select company that was included. I said that this cooler is definitely a step ahead of the Asus and I hope in these coolers will improve things to be improved, ie especially the quality of construction and connection systems. I can say that I saw in a Triton 88 Cooler potential but this was overshadowed by the shortcomings of which I spoke above and which were presented in detail.

Expect from the creators of Rampage series coolers us better quality and a clamping system as solid as the floor's power Rampage 2 Extreme.

Thermalright IFX-14

Surprise test was no doubt this cooler. IFX's probably the most controversial existing cooler because it had many reviews as high-performance and some, including some of us it is not classified as a high performance cooler. "Mystery" was elucidated by testing it on two platforms. While the 775 performed as before (as in my previous tests), the LGA 1366 has proven a very good performance beating even the famous Megashadow at speeds up the environment.

Among the things that led to this reaction is very good on Nehalem better clamping system (I saw what matters is the system for socket 775) and preferably i7 processor's IHS for the bomb (it explained in this article).

I can say that IFX-14 can be a very good choice for those systems based on Intel Core i7 and will use a speed up to 1200rpm. It can also be used without problems on the bench sessions, all if LGA 1366 platform.

Things found in this article reinforces the fact that to draw a conclusion on a cooler, testing must be very complex especially because there are coolers that behave quite differently depending on the test, including the recall and what Z600, that the Nehalem, and especially at a high TDP / lose much of high efficiency. One more reason to emphasize again the need of using a representative test platform for current technology.

PROLIMATECH MEGASHADOW

Tank Taiwanese surprised me several times in this test which defeated the giant Noctua 500 rpm in 1 × 1 × 775 and 800 rpm in 1366, wins very impressive considering its size, you can not compare with his NH-D14 . If May we count on plenty of categories that he was very close to NH-D14 have exceptional image of a cooler, the cooler full to giant Austrian appearance.

He remains a good choice for all conditions of use but especially for low speeds and for those who do not have very much space in case. Overall Megashadow elegant remains one of the best coolers set up in the Inclusive.

Noctua NH-D14

I do not know what I could say now to the end of the article I said before about what was said about this amazing cooler. Things have been pretty clear since the chapter about the "WOW factor", but it was normal to do tests and comparisons in order to draw full conclusions, huh?

Coolant test since I wanted to really defend a cooler impresioant that dominate particular categories of extreme, but I expected too soon. NH-D14 is the most complete and efficient cooler this time. Manages to cool fanless processors with TDP's huge, so impossible to NH-D14, manages to keep under control an Intel Core i7 LGA 1366 to 1.6V with only 1600 rpm, test that you could see in this article. Limits on air cooling under normal conditions, close to the average user have been seriously shaken by this cooler.

I do not think that after I saw these things there are many things to say, I have to say just congratulated the Noctua cooler for creating the best in the history of cooling the air and expect from them based on an extremely cooler tower architecture , as I said in the article "Socket 775 supercoolers.

Finally say that I'd wanted to do more research, because I did everything I proposed and we thank those who helped me and I have been close to achieving this article.

I hope you enjoyed this article and that you found useful, and cool about the test is "likely" that in future we have to test something.

---

There you go. The problem with high end coolers these days is theres really not much of a difference between them especially if you don't go to extremes: fanless or high output fans. Between 1000-1500 rpm its tough to call one or the other better and usually someones choice should be based on price and personal brand preference.

thanks... yeah I lapped my NH-U9B but the concave is supposed to be there - I found out. I want to lap the CPU (now that I know it works) and replace the U9B with a larger, preferrably Noctua so I don't have to replace the mounting hardware.

Overclocked results are in, and have been added to the original post and as well as here.

http://farm5.static.flickr.com/4015/...2e421023_m.jpg http://farm5.static.flickr.com/4011/...b8908cdd_m.jpg http://farm5.static.flickr.com/4043/...6e6ce694_m.jpg

Nice review and pics...but you're missing the big Noctua D14 and the TRUE! :)

These temperatures seems too high for idle processors but too low for 100% load processors to me :S

BTW what is the vcore for your overclock @ 4GHz? :)

Quote:

---

Originally Posted by sniper_sung

These temperatures seems too high for idle processors but too low for 100% load processors to me :S

BTW what is the vcore for your overclock @ 4GHz? :)

---

To the OP: are the graphs showing the Delta? I know it says Air / Core Delta, but normally people show the temperature, and your write-up didn't make clear that you chose to show the [more relevant] change in Delta (change in temperature from idle). I like your way better because ambient room temperature is different for everyone, but I'm still unclear.

Vcore is 1.4 :)

None of the pics in the OP or subsequent posts by OP are showing for me. Just a Flickr message about not being available.

Quote:

---

Originally Posted by bentleya

Vcore is 1.4 :)

---

What's the matter you can't read American? Your customer service is terrible. :clap:

Quote:

---

Originally Posted by SoulsCollective

None of the pics in the OP or subsequent posts by OP are showing for me. Just a Flickr message about not being available.

---

Photos Will be Back up ASAP Flickr is having a hissy fit. Thanks for you understanding.