Intel Core i7 Heatsink Roundup Q309 @ [H]

FrgMstr

FrgMstr

Just Plain Mean
Staff member
Joined
May 18, 1997
Messages
55,634
Intel Core i7 Heatsink Roundup Q309 - Thermalright TRUE processor cooler has been kind of the hill for quite some time. Today we bring in four new challengers to see if any of them have what it takes to make our short list of coolers to buy for your next enthusiast box build.

All the coolers reviewed here today have shown themselves to be up to the task. The question you need to ask yourself is what matters to you? Are you looking for a silent heat sink for an HTPC or the very best heat sink money can buy. Maybe you’re looking more middle-of-the-road for your choice.
Click to expand...
 
Last edited:
Any chance of you guys picking up a dial indicator to measure flatness in the future? Also, how did you guys measure thermal paste thickness to ensure each heat sink was had the same thickness?
 
After applying thermal paste countless times I have simply gotten to the point where I know precisely how much to apply and how to spread it out for a thin, even coverage area.

As far as a dial indicator for flatness, I would say that our temperature readings combined with our grid pattern will show you the level of flatness.

Poor performance mixed with a base that shows warped lines in its reflection is easy to notice.

Regards,
Marc Adams
 
Marc Adams said:
After applying thermal paste countless times I have simply gotten to the point where I know precisely how much to apply and how to spread it out for a thin, even coverage area.

As far as a dial indicator for flatness, I would say that our temperature readings combined with our grid pattern will show you the level of flatness.

Poor performance mixed with a base that shows warped lines in its reflection is easy to notice.

Regards,
Marc Adams
Click to expand...

That's a shame.:(
 
I know that this [HARD] site is meant for enthusiasts and such, but why not test the HSF setups with their provided thermal paste? I know using your own (as stated in the methods) cuts out variables, but it doesn't provide an 'out-of-the-box' review..

On that note, how bad is the provided paste usually?
 
Mazgazine1 said:
On that note, how bad is the provided paste usually?
Click to expand...

It can range from generic white goop to their own brand of thermal paste.

We did a thermal paste roundup and temperatures can range by as much as 2º C if not more using poor quality stuff
LINK HERE

And we feel that most people have their own brand of thermal paste they prefer to use. So we stick with our single TIM.
 
This is what I like to see. No reason to trade in the TRUE, but the next time i need a heat sink for a project I can spend less. It's a win both ways.
 
Marc Adams said:
After applying thermal paste countless times I have simply gotten to the point where I know precisely how much to apply and how to spread it out for a thin, even coverage area.

As far as a dial indicator for flatness, I would say that our temperature readings combined with our grid pattern will show you the level of flatness.

Poor performance mixed with a base that shows warped lines in its reflection is easy to notice.

Regards,
Marc Adams
Click to expand...

Hey Marc, do you ever use the line method when applying paste? Or do you always spread it out evenly over the surface? Arctic Silver says to use the line method, and not spread it out........
 
Because I do not use the AS paste I spread a even layer across the base. This also has the benefit of letting me control how the paste contacts the base of the cooler. Very important to have consistency here.

If I were using the AS paste then I would absolutely follow their instructions.
In The article I linked earlier I tested the AS paste and followed their instructions with good results. You should read that article. It was a lot of fun.
 
There is something I have been wondering for a while, about the orientation of heat-pipe heatsinks. Most of those pipes are designed to use in a mobo setup is flat on table and the pips point straight up and that is how almost every reviews on internet tested.
However, most people install these in a tower case where these motherboards are installed vertically vs ground/gravity. Now these pips are just point side ways and totally defeats the flow of heat(upwards).

Please correct me if I am wrong.

Oh and big typo in article's headline.
"Thermalright TRUE processor cooler has been kind of the hill for quite some time."
 
Last edited:
Trying to find the COGAGE TRUE Spirit on FrozenCPU and NewEgg, but no dice. Any word on where/when?
 
Seems like alot of people areconcerned with how the thermal past was being applied, I am as well.
I kinda paid extra attention to the sections about Base Flatness/Polish. It seems that you favor a highly polished suface. However when you look at the pics of the base after being mounted alot of the heatsinks show signs that they were not making any direct contact with the CPU.
I had always thought that it is best to not to have a highly polished surface so that you dont end up with no direct contact between HS and CPU.

this flash the you linked to in your compound roundup best shows what I am talking about. http://www.mgchemicals.com/techsupport/860_demo.html
 
Last edited:
Hi Marc, great review this time around. May I recommend though that next time you test each heatsink in the round up with 3 separate mounts and runs? That way it can be even more accurate by removing some disparity between mounts and give us a good average to work with. 3 is okay but 5 is preferable IMHO.
 
Man, that Thermalright just won't die. You can call that a best buy since it seems to always be very high at the top no matter the platform. You can probably buy just that one and keep it for years through multiple platforms and still have one of the best performers.
 
I believe the only way to improve the TRUE is to make it a heatpipe contacting surface.
 
Thermalright TRUE processor cooler has been kind of the hill for quite some time.
Click to expand...

However accurate this statement may be (Don't want the Thermalright cooler to think I am saying its anything but KIND) perhaps KING would be the better choice of words in this case...

Kind of sticks out like a sore thumb being at the top of the site and all...
 
Very nice article Marc! Is the loss of a DIMM slot with these heatsinks only a problem withthe Gigabyte X58 Extreme layout, or is it a problem with most X58 boards?
 
visionviper said:
Man, that Thermalright just won't die. You can call that a best buy since it seems to always be very high at the top no matter the platform. You can probably buy just that one and keep it for years through multiple platforms and still have one of the best performers.
Click to expand...

I agree - I have TRUE for ~26 months and I dont expect to change it anytime soon. But it would be nice if someone made even better cooler :)

 
NMR Guy said:
Very nice article Marc! Is the loss of a DIMM slot with these heatsinks only a problem withthe Gigabyte X58 Extreme layout, or is it a problem with most X58 boards?
Click to expand...

It's not just the Gigabyte- I have an Asus P6T and my Noctua mockup covers that first slot as well. However, I have Dominators; if I used standard Corsair DIMMs I wouldn't have a problem with heighth.

Also, Noctua points out here that you can move the fan up a couple of fins if you need greater than the 40mm of clearance.
 
the TRU is back with a vengeance. I thought the TRUE really clobbered the TRU though :confused:
 
Thanks for the roundup.
I have an Antec Fusion Remote Max HTPC with a Gigabyte EX58-UD4P mobo. How can I make sure which of these heatsinks would fit? Some heatsinks hang over one side more than others.
 
wolfhero said:
There is something I have been wondering for a while, about the orientation of heat-pipe heatsinks. Most of those pipes are designed to use in a mobo setup is flat on table and the pips point straight up and that is how almost every reviews on internet tested.
However, most people install these in a tower case where these motherboards are installed vertically vs ground/gravity. Now these pips are just point side ways and totally defeats the flow of heat(upwards).

Please correct me if I am wrong.

Oh and big type in article's headline.
"Thermalright TRUE processor cooler has been kind of the hill for quite some time."
Click to expand...
I understand completely where your coming from, because i was very confused by heatpipes for the longest time. Heatpipes actually get very little from the thermosiphon effect you are thinking off. Thermosiphoning is the natural heat convection cooling that can be used in a closed loop and create a flow. Thermosiphoning ONLY WORKS IN THE PRESENCE OF GRAVITY. It's based on "hot goes up" and "cold goes down".

Heatpipes are pretty much completely powered by capillary action. The heatpipe is internally covered with a "wicking" material and a hollow space. when any part of the heatpipe gets hot, the small amount of liquid in the pipe evaporates from that surface, and condenses on cooler parts of the heatpipe.


Meanwhile, the hot spot that is now dry is fed with more liquid by capillary action from nearby wicking material.

The cycle continues.

I have read that heatpipes do indeed benefit in certain orientations because of the thermosiphon effect, but the difference is small.

http://en.wikipedia.org/wiki/Heat_pipe

edit: that entire post was kind off topic... so back on topic. COGAGE TRUE Spirit looks absolutely awesome. I remember i paid 40 bucks for my slk800 thermalright. Can't imagine paying the 60 bucks for most new air coolers.
 
Shlomo said:
Hi Marc, great review this time around. May I recommend though that next time you test each heatsink in the round up with 3 separate mounts and runs? That way it can be even more accurate by removing some disparity between mounts and give us a good average to work with. 3 is okay but 5 is preferable IMHO.
Click to expand...

I think this would be a good improvement to the methodology. Although it would be a PITA to actually have to do it. Haha. Also, I wouldn't re-mount the same heat-sink three times in a row, assuming you had three heatsinks, I would rotate them as suc:h 1-2-3-1-2-3-1-2-3 I think this would further eliminate (through averaging) any inconsistencies that could sneak into the data.
 
I'm genuinely shocked that the load tems are so low (even for the stock fan)

I currently have my i7 920 at 4400 (210x21) @ 1.35v and I routinely hit low 80s under full load. I use the AVC ESP Cooler, which in other reviews, ranked just below the Noctua.

Are other people getting temps this low? I was under the impression that overclocked i7s routinely go in to the 70s and 80s when stress testing.
 
Shlomo said:
Hi Marc, great review this time around. May I recommend though that next time you test each heatsink in the round up with 3 separate mounts and runs? That way it can be even more accurate by removing some disparity between mounts and give us a good average to work with. 3 is okay but 5 is preferable IMHO.
Click to expand...

I'll second this. I've already stated my fears about uniform TMI applications and I think this would alleviate it at least to some degree.
 
I think the fact that it's an open setup has a lot to do with the lower-than-average temperatures. The air being brought into the heatsink is always ambient temperature (25C), which is different than in a case-contained setup, where the hot air is often retained in the case (to varying extents depending on air flow, but it will never be as efficient as an open case set up in a large room.

The open set up also minimizes the effect of heat emanating from other components like the power supply and the graphics card. Further bolstering this theory is that the graphics card used in the setup was selected due to its low heat footprint. Most users on this forum pair their OC'ed CPU with a graphics card that is performance-driven, and puts out heat on par with that performance (especially if the card's fan directs some air back into the case).
 
The Cogage looks just like the Sunbeam cooler I have been using for years successfully. Is that just a renamed product?
 
So do any of these actually beat the IFX-14 that you can pick up for $60 now at a few places? The IFX-14 had the best cooling from what I saw.
 
0xygenthief said:
However accurate this statement may be (Don't want the Thermalright cooler to think I am saying its anything but KIND) perhaps KING would be the better choice of words in this case...

Kind of sticks out like a sore thumb being at the top of the site and all...
Click to expand...

Yeah it's rather distracting, being in the first sentence of the headline and all...:eek:
 
thanks again as usual, great review and great comparisons w/ what's out on the market!
 
anthrex said:
does anyone know how well the Thermalright ultima 90 cools the the 920?
Click to expand...

It would probably struggle w/o a high speed fan, IIRC the Ultima-90 was one notch below coolers like the Xigmatek S1238 and others which were basically one or two notches below the original TRUE... Depends on the fan and your level of OC, shouldn't be a problem at stock speeds though. Just a guesstimate on my part...
 
Man, this is making it very hard for me to decide between the Noctua with a single fan or the Megahalems. There are worse problems to have, both are amazing coolers that produce minimal noise.

I dunno, maybe the Megahalems with a Noctua fan attached. :D
 
5 degrees difference when OC'ed at my speeds of OC ain't that much to make me drop my Noctua to Cogage... though it is tempting. But I prefer to trade a bit of heat for lack of noise ;)

Now, found something interested in Legion review: "When installing the TRUE Spirit we noticed that the clips required a serious amount of force to be installed correctly. The problem with this is that we noticed that the heatsink did tend to bend the motherboard quite aggressively, which could lead to problems."

Did you guys run into that problem with Cogage too? Or it's just you know how to properly mount it? :D
 
It has the same mounting as an Intel stock cooler... which is good and bad...

I still prefer bolt-thru kits...

I have a TRUE120 but it still doesn't feel solid on the CPU... you can wiggle it with enough force, My Noctua never moved, nor my Xigmatek. I'm not sure if this is because of the non-flat base or wut. I did the washer mod too
 
You must log in or register to reply here.
Back
Top