Mega-Heatsink Roundup - 0407 @ [H]

One more vote for Thermalright here :) I think the Ultra 120 is a great HS period.

I also love heatsink reviews. I have a Thermalright HR-03 Plus on my 8800gtx and saw drops of almost 30 degrees from stock cooling :p You should do a 8800 heatsink review next ;) I think there may only be 2-3 coolers though..haha
 
My temps with ThermalRight Ultra 120 Extreme with 2 fans (1200RPM Scythe S-Flex push-pull config) after running 24 hours F@H - Q6600 (B3) @3.4Ghz Vcore1.4

 
I also forgot to address a point : Is all tests done with just 1 mount per test ? The reason is because sometime, doing 3 mounts will help give a more accurate result and chalk off variations due to a possible bad mounting.

 
I have to agree that you should have used a quadcore.

I've seen other reviews where some of those heatsinks become heatsoaked at higher themal loads, again using a overclocked quadcore.

Speaking of which. I have a QX6700 that I do not need.

I am willing to part with it for a mad silly cost of $150 shipped for [H]ardOCP only.

Kyle or Marc PM me if you're interested.
 
Interesting review with a little novelty thrown in. Not picking a "winner", value estimations and "best looking" are fun. I'd also have loved to see a quad core used and I do wish you'd included The Freezer 7, It might not have threatened the top rank performers, but I gotta believe it would have run away with the vaule ranking...
 
Mackintire said:
I have to agree that you should have used a quadcore.

I've seen other reviews where some of those heatsinks become heatsoaked at higher themal loads, again using a overclocked quadcore.

Speaking of which. I have a QX6700 that I do not need.

I am willing to part with it for a mad silly cost of $150 shipped for [H]ardOCP only.

Kyle or Marc PM me if you're interested.
Click to expand...

With the new processors moving about, we should have an extra for the next project. Sometimes equipment is not easy to furnish to everyone that needs it at the time. :( I found a 3.8 Prescott last week, maybe we should use it. :)
 
The only reason I figured that QX6700 would be ideal was because it does not throttle till you hit 100 C. That makes the QX6700 a little more interesting as a test CPU.

But, given what you said I understand.
 
I'm a little dissapointed that none of the test where conducted fanless. I can run my Athlon 64 x2 4400+ at 100% utilization with a Sythe Ninja and no fan at stock speeds while staying in manufacturer tolerances for thermals. Works fine for games too.
Would be nice to see which heat sinks can be used without a fan for those of us who run home theater boxes or are silence nuts.
It would also be an excellent way to actually show how well the design of the heatsink is. Fans add alot of variables.
 
Great article, I like the "no nonsense" approach to just posting the graph and worrying about the details later. Not that the details aren't important, I just like instant gratification :p.

That said, I'm a little disappointed that I never see the "Enzotech Ultra-X" in these "shootouts". My 6600 Dual Core runs at 3.35ghz in a fairly hot environment (average temp ~78 degrees) with at least one core at 100% 24/7, and with the Enzotech my core temps are rarely over 45 degrees C, usually more like 40-42. What's unique about it is that the heatpipes are 8mm in diameter, vs. the more common 6mm. See the link below for a review on it with more details. It would have been nice to see how it stacks up against the more popular heat sinks in this test.

Enzotech review: http://www.frostytech.com/articleview.cfm?articleID=2082
 
ultra120 review said:
To install the heatsink, you will have to remove the motherboard which is never fun. In our case here the heatsink was mounted in a different position due to the tall cooler on the Northbridge. If you have a similar setup thank you can expect the same issues with your installation.
Click to expand...

I'm confused
 
Mackintire said:
I have to agree that you should have used a quadcore.

I've seen other reviews where some of those heatsinks become heatsoaked at higher themal loads, again using a overclocked quadcore.

Speaking of which. I have a QX6700 that I do not need.

I am willing to part with it for a mad silly cost of $150 shipped for [H]ardOCP only.

Kyle or Marc PM me if you're interested.
Click to expand...

That's true... I tried running 3.6 GHz @ 1.45v on my Q6600 and my Thermalright SI-128 became so overwhelmed it just kept rising and rising slowly till it reach 95C before shutdown. I tried running the fan at max speed and aimed another fan close to get rid of the heat without success. I guess the heatpipe coolers will reach a point where it cannot dissipate the heat fast enough to work.

I hope Kyle will find a quad for the next round.
 
Wow, an [H] Heatsink roundup, I can't even recall if there ever was one before :)

Couple of points I disagree with.

1. If the heatsink you are testing comes with a fan, use it, that will truely show the value for the $, and those that don't come with one should have that extra cost added to their final cost.

2. Noise testing, get a good sound meter, it is allways nice to know which coolers can do what for how loud they are, which can be a make or break buy for a lot of people.

Other than that, great review, unfortunately I'm on water so it doesn't apply much to me.

Maybe a true water cooling kit roundup would be a good idea in the future, tho expensive if you have to buy them all.
 
What difference does it make if you test with a quad? Enthusiasts are not all going to run the same CPU anyway, so it isn't necessary to standardize testing on a particular (type of) CPU. Three months down the road, everyone will have moved on to the next chip.

Instead you could build a test rig with a hot plate, a variac, thermocouples, and sound meter. You could calibrate it with one of those Kill-a-Watt power meters.

Was 0407 really supposed to be Q407? Cmon, you can tell us...
 
Did I miss it or you never specified what were you using to measure the CPU temp? :confused:

EDIT: I think 0407 means "April 2007"
ED2: Nevermind, I found it. Nevertheless, I think you should have used CoreTemp 0.95.4.
 
Instead you could build a test rig with a hot plate, a variac, thermocouples, and sound meter. You could calibrate it with one of those Kill-a-Watt power meters.
Click to expand...

Another good idea.


FYI the QX6700 can exceed 260 watts of heat when overclocked. I 'll still stand by it being the hottest CPU intel has ever made. The 3.8 prescott is another rediculus hot beast, and I would say that the 3.8 prescott is even hotter at default speeds, but when overclocked the QX6700 beats it and can take down a fair share of the heatsinks on the market. Full load on a Scythe Ninja version A with an 70 CFM fan running 3.2Ghz I was averaging 85 C load. The noctura UH-12 which they reviewed here dropped it to 74C

My new QX6850 running the same clock speed is 72C at full load at default voltage. Much better.

Part of the reason the Ninja didn't cool very well was because the CPU was emitting more heat then the heatsink could deal with....regardless of the fan.

I 've read that at about 3.4Ghz with the QX6700 my noctura will go into themal runaway and that only 2 or 3 of the heatsinks reviewed here could cope with that kind of heat.

I've never run my CPU that high so I don't know from personal experience. That and I like to run EIST.
 
I was hoping to see how much power the TEC coolers use as they will be drawing power from the PSU.
Otherwise its a really good review, I didnt know some of those heatsinks existed.
 
I liked the review however as others have noted a few things could have made it perfect...

It would have been nice to see a quad Intel reviewed. Having a Q6600, I found out in a big hurry how fast a quad can saturate the heatsink. It would have been nice to see the comparison on it as well.

The price/performance is not entirely accurate from what I could tell because even though the TR Ultra 120E won that comparison, it does not come with a fan. This implies a better value than is actually there.

It would have been nice to see the coolers reviewed with the fan they came with along with the "control" fan. It may have been better to use those type of results for the price/performance ratio (factoring in cost of a fan to a sink that didn't include one).


Now to contribute to this post...

I have the TR Ultra 120E and as mentioned in the review the convex base can be a big problem. I even seen reports of users saying TS claimed that it was a "feature" because the heatspreader was usually concave. But, in my case the base was so convex there was less than a 1/4" dot that was actually in contact with the cpu. The thing rocked and twisted an "extreme" amount (pun somewhat intended). Since I wanted to get the build done and didn't want to mess with returning it and waiting for a replacement, I wet-sanded the base to a flat surface with 220/400/600/1500 (which was an additional expense). Once that was done it seated very well, did not rock or twist, and performed as expected.

I first mounted the unit to exhast to the rear of the case, however in my setup I wasn't getting a good overall system cooling. I then mounted the unit to blow upwards in the case, which made a massive difference in the overall cooling of the NB & SB by expelling the hot air around the video card (but with the negative effect of moving warmer air thru the sink, however in my case the trade-off was well worth it). I did find that the airflow of the S-Flex SFF21F I originally bought for the setup just didn't cut it for my use. I picked up a higher CFM Sunon that was a bit louder and bigger but it made all the difference. All said and done, I really am happy with the Ultra 120E and would recommend it to anyone (that is if they are not afraid of returning or sanding the base if it is not flat). Just make sure that it can fit in your case!
 
Great review, thanks. Some of the pictures don't work though like the SCYTHE INFINITY and THERMALRIGHT ULTRA-120 EXTREME.
 
Marc Adams said:
Thank you to everyone who has taken the time to read the article. I certainly hope you enjoyed reading it as much as I did preparing it for you.

Like any great piece of work there will be revisions, tweaks and adjustments so keep those comments (good & bad) coming. Where else can you be certain to have a hand in the development process of upcoming reviews!




With so many different cooling products out there and so many on the way you can be sure there will be more articles to come! :D
Click to expand...

Nice article :)

I hope to see a Zalman in the mix next time :)
 
good baseline guys, good write-up, clear charts, focusing on performance is a valid approach.

But what about noise? :) That Freezone unit is very very loud; Kyle, buy this guy a dBA meter!
 
Nice, and this unbiased review was a surprise. Not that I ever doubted the [H], more so that I just read some pretty slanted reviews from 2 other large hardware reviewers.
 
First question to [H]: did you notice at all any movement with the Thermalright?

You said it seemed to make good contact, but we (or at least I) need to know if that's with one of 'em that wasn't turning all over the place.

Does is make good contact even when it can be turned far too easily?
Or did you guys happen to get one that was actually flat and sat sturdier?

As for the Zalmans, I can give a comparison, considering that's all I've used for years, but I still have an issue with the Thermalright Extreme vs something like the Zalman.

I use the Zalman 9700 on my C2D E6700 and my temps are better than than Thermalright by at least 1-4 degrees C both idle and load. This is in relatively hot heat as well and with the Zalman's fan speed set to low, and it's not even a 120mm fan. I idle at 30C at the highest and hit 41C at the very highest as far as I've ever seen. Granted, I do have two 120mm fans for exhaust and four 120mm side case fans (NZXT case), but I'm also running an 8800GTX and my PC is on at least 12 hours or so out of a day (work and gaming).

I have the Thermalright, but after test mounting it on one of my boards and seeing how much/easily it turns, I put it back in the box and in my closet in case I ever decided to try it again.

BUT... I do have an interesting issue: with fan speed set to low, the Zalman has a very annoying "wobbly" noise that I can hear over all the 120mm fans in my case, which on their own are extremely quiet believe it or not and my system can barely be heard.

Now, the Zalman has better temps than the Thermalright (take note of that, for those who want to know) with a 92mm fan on slow speed. With my set-up, anyway. But, that "wobbly" noise is finally getting to me after all this time, and I've thought about moving to a tower style cooler so I can use a quieter 120mm fan.

I am concerned with all the movement with the Thermalright that it wont make/keep good contact. I'm one who believes something should work as intended out of the box, and I'm not going to start shoving cardboard, pennies etc. under the base of it just to keep it from moving. That's shoddy workmanship and a shoddy product if you have to start doing "construction" on it right out of the box.

I'm also of the mind that, "if it ain't broke, don't fix it", and with the better temps on the Zalman, I'd just leave it be if the damn "wobbly" noise wasn't starting to make me homicidal.

However, the other couple of tower designs reviews did not do as well as the Thermalright and, me being an extremely heavy gamer and my PC on at least 14 hours out of a day, temp is important.

So, do I chance it?
Did [H] find the Thermalright was moving/could be moved too easily but still gave those temps?

It would allow for me to use a 120mm, which will be quieter like the rest in my case, as well as it being manually speed controlled, as I'm using on my GTX's cooler. But I don't want to go through the hassle right now if it's going to end up being screwed. It's hard for me to tell if my Thermalright's base is not totally flat.

The Zalman speed controllers never work (keep that in mind) and my bios has two system speeds for fans: low and high. On high, the Zalman literally sounds like a damn jet about to take off! Haha.

But since it does better than anything else except water cooling (even shown if I compare to others in the review of coolers here) from my experience, with the fan on low, it's not a problem. But if I want to use a speed controlled fan for temp and sound control, I need a hsf that will allow for use of any fan I can plug into my PSU, 4-prong that is speed controlled.

That would make use of the Thermalright worth it IF the damn thing would actually make good contact and not move all over the place.

There seem to be quite a few using the Thermalright here.
Are you all having it move on you? How is the contact/temps?

Anyway, that's my "rant" of some specs and questions.
I'm ready to tear down and try to slap it in right now, but on my test mounting, I screwed it in so tight I could have cracked my board, and the thing still moves way too much, IMO.
I didn't check thermal paste coverage though to see the contact at that time.
Not sure what to do.

If it weren't for the damn "wobbly" noise of the Zalman, I wouldn't even think twice. But, it's starting to drive me absolutely nuts and I'm in front of my computer many hours during a day.

Any feedback welcome.
Off to go look at the damn Thermalright again and contemplate :rolleyes:
 
Bdbtoys said:
I liked the review however as others have noted a few things could have made it perfect...

It would have been nice to see a quad Intel reviewed. Having a Q6600, I found out in a big hurry how fast a quad can saturate the heatsink. It would have been nice to see the comparison on it as well.

The price/performance is not entirely accurate from what I could tell because even though the TR Ultra 120E won that comparison, it does not come with a fan. This implies a better value than is actually there.

It would have been nice to see the coolers reviewed with the fan they came with along with the "control" fan. It may have been better to use those type of results for the price/performance ratio (factoring in cost of a fan to a sink that didn't include one).
Click to expand...

Again great feedback! Keep it coming. We will make these adjustments.

1. Quad Core
2. Stock Fan used
3. Add fan cost to value graphs
 
JMke said:
good baseline guys, good write-up, clear charts, focusing on performance is a valid approach.

But what about noise? :) That Freezone unit is very very loud; Kyle, buy this guy a dBA meter!
Click to expand...

Yep, dBA meter is easy to do now days. Radio Shack has some good ones.

Marc, does your setup allow for noise levels to be taken in a static manner so we could get comparable results from say 6 feet?
 
ChthonicEidolon said:
Any feedback welcome.
Off to go look at the damn Thermalright again and contemplate :rolleyes:
Click to expand...

My feedback.....and I have been a huge Thermalright fan for years. We have used their products for years and worked with them in the USA before they were known and I have never had issues with their products. But given what I have seen about the mod that need to be done to some of these units, I would suggest scratching them off you list if you have worries. The fact is that there are too many great units out there to chance getting a bad one by an unlucky draw. Go with another and be happy! :)
 
To Eidolon: If your Zalmann is getting better temps then the Extreme, then you need to lap your Extreme. Sad but true.



For what it's worth, I run the Gemini with 2 Antec Pro120mm fans on my lapped 6300. At 3.4GHz @ 1.485v it never sees 60C running Orthos blend and TAT will peg it to almost 70C. Thats with a 24C room temp and case temp in the high 30s. The 2nd fan makes a difference, and so does a flat heat spreader. That being said, it does have a limited capacity, a quad would saturate it pretty quick.

Take that line on the test page about TAT being hotter then anything else to heart peeps. TAT results will never be duplicated running real software, even if it loads both cores.

dBA meter for sure, as has been mentioned.

Drat, Kyle just made most of my feedback moot :)

In that case just a question. Is it just me, or does that thermal paste look a lot like the stock Intel stuff you get in the syringes (not sure if the public sees these) because I love that stuff. NO cure time, great performance, and stupidly easy clean up. It feel more like a clay then a paste.

Oh, and if that other guy quit, is Marc (good job man) just filling in? Y'all need a new cooling guy?

Nice read, thanks.
 
On page 11 for thermalright's heatsink I think you meant Thermalright is one of the few competitors... No big deal.

I like the heatsink reviews a lot, tells you what you are going to be getting for the money.
 
Very nice review, just got around to reading it. Very big sample.

I'd love to see some ready to use watercooling kits compared to the air coolers you had.

I'm interested to see how my Nautilus 500 would stack up.
This must have been alot of work. Great job.

How about ArcticCooling products as well next time.:D
 
magoo said:
Very nice review, just got around to reading it. Very big sample.

I'd love to see some ready to use watercooling kits compared to the air coolers you had.

I'm interested to see how my Nautilus 500 would stack up.
This must have been alot of work. Great job.

How about ArcticCooling products as well next time.:D
Click to expand...

Agreed, I'd like to see more of those preconfigured watercooling kits similar to the nautilus and the coolit ones.

I'd also like to see a nice GPU heatsink roundup :)
 
nice article, but for me, ultimately, it doesn't have much meaning without providing noise levels.

If the reader is just looking for the best cooling or best cooling for the money, this article has what they need (though they might want to see how the Gemini performs with the extra fan), but when I buy aftermarket coolers, noise is huge factor.

I definitely want better performance than stock (and above average overall), but I also want to achieve that with minimal noise.

Then again, I guess that's what SPCR is for.

I still love you guys ;)
 
Long time reader first time commenter, loved the article, reminds me of the [H] of old. Any chance you guys can do a Heatsink Roundup that fits in SFF type cases like the X-QPACK *hint hint* ;) Finding a great cooler with limited head room has been an issue for me and others I've talked to with the case.
 
Great review guys! I've been hoping for a recent cooler round-up and its awsome to see how the air units perform against mtec...

A lot of good feedback. I am however surprised that no-one brought this up: What about the newer HDT (Heatpipe Direct-Touch) coolers? I'm a bit surprised that a modern round-up doesn't include the most recent cooling technology for air-cooled units. If you'd like I can dig up a few recommendations for your next round-up. I have not yet seen a review that covers them next to the best coolers you had in your round-up. The best review I've seen with the HDT coolers compares them with the Zalman 9700 which is ancient and I really don't even know how that one compares to the TRUE and the Tuniq.

I know it'd be a lot of work but it'd be nice to see a collection of coolers and their performance in charts, kinda like TomsHardware's charts but for coolers { and unbiased :) }

Thanks again and keep up the good work!
 
This was a great article. Logical layout, good methodology, nice analysis, great open-ended conclusions.

Unfortunately, both the author and the editor managed to miss several instances where an apostrophe was used for pluralization instead of possession. This is a fourth grade grammar error, and unfortunately it's becoming more and more common. I just don't want to see it become "accepted" because so many people do it.

I'm being a little bit of a grammar fascist, I know. I'm not taking anything away from the great work you guys have gone on this article-- it truly is great work in all other ways. But seriously, if you're going to be in the product review game and be taken seriously, the grammar has to pass muster.

Just trying to keep you honest, sorry to be a whiny b#@%^. ;)
 
Just wanted to follow up with those HDT coolers.

OCZ Vendetta http://benchmarkreviews.com/index.p...ask=view&id=90&Itemid=62&limit=1&limitstart=0

Zaward VIVO http://benchmarkreviews.com/index.p...ask=view&id=92&Itemid=47&limit=1&limitstart=1

Xigmatek HDT-S1283 http://benchmarkreviews.com/index.php?option=com_content&task=view&id=81&Itemid=47

A page that compares the 3 of them and a few other coolers: http://benchmarkreviews.com/index.p...ask=view&id=90&Itemid=62&limit=1&limitstart=4

Again, I'd love to see how these coolers (this technology) compares with the TRUE and the Tuniq Tower.

Oh, and it was the Zalman8700 they used...
 
sublifer said:
Just wanted to follow up with those HDT coolers.

OCZ Vendetta http://benchmarkreviews.com/index.p...ask=view&id=90&Itemid=62&limit=1&limitstart=0

Zaward VIVO http://benchmarkreviews.com/index.p...ask=view&id=92&Itemid=47&limit=1&limitstart=1

Xigmatek HDT-S1283 http://benchmarkreviews.com/index.php?option=com_content&task=view&id=81&Itemid=47

A page that compares the 3 of them and a few other coolers: http://benchmarkreviews.com/index.p...ask=view&id=90&Itemid=62&limit=1&limitstart=4

Again, I'd love to see how these coolers (this technology) compares with the TRUE and the Tuniq Tower.

Oh, and it was the Zalman8700 they used...
Click to expand...

Wow the Xigmatek HDT-S1283 ROCKS!!!
 
You must log in or register to reply here.
Back
Top