• Welcome to Overclockers Forums! Join us to reply in threads, receive reduced ads, and to customize your site experience!

A few questions about heatsinks...

Overclockers is supported by our readers. When you click a link to make a purchase, we may earn a commission. Learn More.

Aynjell

Member
Joined
Dec 12, 2008
With the rising thermal load produced by processors, I'm finding it really hard to select a heat sink for an i7 processor I intend to buy. When I bought my 3800X2, the clear choice was the Zalman CNPS9500, it performed similarly to the rest of the market, looked bad-***, and wasn't too overpriced.

now we're getting into coolers that weigh anywhere from 700 to 1000g, and require push pull fan situations just to keep the CPU under 70 degrees Celsius when overclocking. Coolers are getting so big that getting solid copper units is ill advised due to load on the motherboard. (for example, the copper true is 1900g, almost twice the already heavy coolers, and for reference, my current CNPS9500 is 530g, and it's all copper) So, on to my question:

The heat sink market is a very confusing one, and unfortunately, no amount of reading reviews appears to be clearing it up for me. The heat sinks I've all seen reviewed show the coolers reviewed with the stock fans they are shipped with, and I am convinced these fans are not ideal as they are usually the manufacturer's in house unit. Two heat sinks I'm considering are the Noctua NH-U12P and the Thermalright True (black), both with a push-pull fan setups. The reviews I've read show these units with stock fans, and the Noctua won the round up. The true is a favourite amongst current generation overclockers, and I know it has to be for a reason. I know the Noctua is also held in high esteem and it performs well at it's price point, seemingly providing better value than the Thermalright.

Here's my questions:

Which heat sink works best after aftermarket fans are installed?

Also, with the increase in weight on upper grade heat sinks, have you all noticed any stress to the motherboard?
 
I dunno if it helps but we have here a test in open stand with plain TRUE, black TRUE, copper TRUE, CM Z600 and TR HR-01 Plus with a 3000 RPM fan. CPU was Quad Q6600 G0 @3711mhz - 1.504/1.488 Vcore idle/load and 21°C room temperature.

attachment.php


IMO you can't go wrong with TRUE, Noctua, Mugen2 or CM Z600.
 
Hrm, half a temp more on the black, so getting the true which costs a bit less might be a better choice. I am a mustang owner and have had the mentality built into me that sacrificing anything performance wise for aesthetic is rice. :(

I mean, half a temp is nothing, that's like, margin of error type stuff. That could be a bad application of as5 for all we know or basic cooler issues. Obviously though, the copper version is a bad deal. it weighs and costs way more without any advantage. twice the weight, damned near twice the cost, with no benefit in that test. I can't click the link at work, google translator is blocked. What kind of fan were they using?
 
Burebista and/or others,

Is it helpful to lap a TRUE/heatsink, for example, and NOT lap the CPU? I may be getting a TRUE and was wondering if it a lap it all or nothing type situation? I am more concerned about my ability to resell my cpu down the road and maybe lapping the cpu would prevent it from selling. I am curious.
 
Sorry man, I'm not a clocker maybe others can help you. :)
As you see, that friend of mine lapped all those TR heatsinks. Only plain TRUE and copper benefits from lapping, the others two not a single °C so YMMV.
 
Aighty. Seems like either way I'm stuck between 60 and 70 with even a really good case. I may be looking at water cooling then.
 
Burebista and/or others,

Is it helpful to lap a TRUE/heatsink, for example, and NOT lap the CPU? I may be getting a TRUE and was wondering if it a lap it all or nothing type situation? I am more concerned about my ability to resell my cpu down the road and maybe lapping the cpu would prevent it from selling. I am curious.

Yes, it is helpful to lap the TRUE even without lapping the cpu. The TRUE usually has a definite convex shape to the base from the factory and doesn't make good contact without lapping. Depending on the flatness of the cpu heat spreader, lapping that may may not drop temps some more from just lapping the TRUE but I think the biggest improvement will come from lapping the TRUE itself. I've checked my last two AM2 processors and they are very flat from the factory. I knocked off about 4 deg. C under load from lapping just the base of the TRUE.
 
Yes, it is helpful to lap the TRUE even without lapping the cpu. The TRUE usually has a definite convex shape to the base from the factory and doesn't make good contact without lapping. Depending on the flatness of the cpu heat spreader, lapping that may may not drop temps some more from just lapping the TRUE but I think the biggest improvement will come from lapping the TRUE itself. I've checked my last two AM2 processors and they are very flat from the factory. I knocked off about 4 deg. C under load from lapping just the base of the TRUE.


Thank you Trents for your helpful response.
 
Back