Intel Core i7 Cooler Review @ [H]

FrgMstr

FrgMstr

Just Plain Mean
Staff member
Joined
May 18, 1997
Messages
55,601
Intel Core i7 Cooler Review - New CPU coolers from Kingwin, Thermal Transtech International, and ZALMAN. While air cooling has gotten about as good as it can possibly get, all of these coolers bring something unique to the table. But do these units perform? We find out.
 
Surprising results with the Zalman 9900. I used one on my Phenom II before I got the Megahalems, and even at full speed, it kept my CPU 5-6 degrees higher than the Megahalems and a 1600rpm Scythe Slipstream. I wander if it's because the AMD mounting system is total crap. Nice write-up BTW.

Edit: Went through my old logs from my cooler testing and I updated the post with accurate information.
 
Last edited:
Surprised the Zalman did that well, thought it was going to be another under par pretty looking heatsink. I want to know where I can get a 79500 gt ;).
 
nPower Hurricane is interesting. I wonder what its pricing will be? Even with one fan it was running great.

I've been eying a few HSFs , any opinions on the Scythe Mugen 2 Heatsink? Its really really cheap compared to others, and its received positive reviews.

P.S. - I didn't notice the Apples to Apples section before, very nifty.
 
Something is wrong with the data in the price to performance ratios for the TRUE spirit and the TRUE-120. I noticed because the TRUE-120 should have a near equal ratio to the Megahelems from the price and performance but the TRUE-120 is given a better number. I decided to run the numbers after that.

TRUE Spirit
37.99/26.9=1.41

TRUE-120
59.99/23.7=2.53
 
What a great Review! Now if we can convence Intel to do away with the Stupid Push Pins and adopt something better....

Hopefully the Heat Sink Manuf will list the MB's it fits on. ie no Suprises.

Since my next build is over a year off I plan on waiting (E8400, 8g Ram).

DM
 
So the TRUE Spirit is about as good as the T.R.U.E. for a lot less and in includes a fan. Sucks for the people who bought the ol T.R.U.E.... oh waits thats me!
 
Jakalwarrior said:
So the TRUE Spirit is about as good as the T.R.U.E. for a lot less and in includes a fan. Sucks for the people who bought the ol T.R.U.E.... oh waits thats me!
Click to expand...


And on that note, why is the TRUE spirit three degrees different from the TRUE-120 in the cost to performance table? It was never more than a degree away in the performance charts.
 
I really do wish there was one GIANT CHART with all the previously tested 1366 coolers on it. It's nice that they reference the stock cooler and the TRUE's but I can't help but think keeping all the heatsink roundup's in one chart would really make selecting a heatsink easier.
 
I'm using Thermalright Ultra 120 eXtreme:D


123.jpg
 
Can a Thermalright 1156 bolt thru kit be used with the Cogage True Spirit? A google search came up with a hit that suggests you could, but I haven't found a definitive yes.
 
I wonder how much airflow the Zalman is losing radially from the fan since it is lacking a shroud.
 
Crispy002 said:
I really do wish there was one GIANT CHART with all the previously tested 1366 coolers on it. It's nice that they reference the stock cooler and the TRUE's but I can't help but think keeping all the heatsink roundup's in one chart would really make selecting a heatsink easier.
Click to expand...

That most likely wouldn't be accurate though. The test setup would have to be identical. Ambient room temperature, case air flow, specific processor used in the test bed, etc would all have to be the exact same. You can be as meticulous in your testing procedures as you want but there's always going to be some factor that's a little different when you don't test them all at the same time. I doubt many hardware review sites have the capabilities to do a 30+ heatsink round up in a single day on a single test bed.
 
I wonder how much airflow the Zalman is losing radially from the fan since it is lacking a shroud.
Click to expand...

Initially the CNPS9900 did ship with a shroud, but even then Zalman recommended the removal of it. It increased both the temperature and noise level. To be honest I think the staggered orb shape is more for looks than function.

I'm guessing there's no plans for a Phenom II cooler roundup?
Click to expand...
Most high-end coolers ship with the necessary components for S775 and AM2/3 mounting as well. Just keep in mind that the AMD socket is rotated 90 degrees compared to Intel so the cooler might have to face upwards instead of blowing the air toward the rear exhaust fans.

Surprising results with the Zalman 9900. I used one on my Phenom II before I got the Megahalems, and even at full speed, it kept my CPU 5-6 degrees higher than the Megahalems and a 1600rpm Scythe Slipstream.
Click to expand...

This surprised me too. I'm using this heatsink now and while it's better than the Auras92 I used before (by 12C or more, in fact), it's not the greatest AMD cooler, especially considering the price. It's however extremely easy to mount on AMD boards and can be oriented in the horizontal position which is great if your case lacks a top exhaust fan. Maybe it's better with Intel CPUs, although that doesn't make sense either...
 
Thanks for the cooler review, [H]. :)

I'm using ThermalRight Ultra Extreme 120 Black and see no reason for me to upgrade from what I currently have. The cooling performance for my system is outstanding. To only gain a couple of degrees increase wouldn't be worth it to me cost wise to switch.

I will recommend the Zalman ZNPS9900LED to people building new systems since most of my friends are overclockers.
 
This means drilling a very small path into an expensive CPU to place our thermocouple.
Click to expand...

drilling <-------milling

What can I say, im a machinist some times ;)
 
Still no IFX-14 in these reviews... what, does everyone fear the king?

I find the results of the Zalman odd... almost 'please welcome our new sponsor, ________' odd.... lol. But seriously...
http://www.bit-tech.net/hardware/cooling/2009/02/19/ga-1366-cpu-cooler-group-test/7

http://www.xbitlabs.com/articles/coolers/display/core-i7-coolers-roundup_18.html

http://www.frostytech.com/articleview.cfm?articleid=2362&page=3

Something tells me that the degree of control in this experiment was not up to snuff. In most other tests, the 9900 doesnt even come close to top dog.
 
Last edited:
I'd assume the top bar and bottom bars on the Sound Chart are the Idle and Load temperatures respectively?

Its interesting to see such a big difference between Idle and Load temps on the Zalman and Kingwin.
 
Since the Kingwin doesn't support 1156, maybe you should rename the article the LGA 1366 cooler review.

Just nitpicking :p
 
This is a great review. I might be doing a build for a friend's sister. I probably would have just gone with a TRUE 120 but now I'm kicking around the Zalman reviewd in here instead.

Oh, saw a type on page one, the introduction page:

"For good measure we include ZALMAN&#8217;s newt cooler the CNPS9900 LED..."

(Boldface and italics my addition).

I just have to ask..would the Zalman also work on let's say a salamander? :D

Fixed! Thanks for the extra eyes. Kyle
 
Last edited by a moderator:
these are all too loud =\

i wish they'd give us more info on the sound aspect. some fans make high pitched noises and others make lower less annoying sounds. just posting DBs isn't enough.
 
martin_nj said:
these are all too loud =\

i wish they'd give us more info on the sound aspect. some fans make high pitched noises and others make lower less annoying sounds. just posting DBs isn't enough.
Click to expand...

Well, the pitch is dependent on the speed of the fan. Any cooler will make a high pitched racket with the fan at High. However if you have a very good cooler, you can run the fan at a lower speed, say 1400 RPM and it will be quieter and less annoying. A poor cooler might require the fan to spin at 2500 RPM to be effective.

I prefer that coolers ship with a loud, high performance fan. It's easier to slow down a loud fan via e.g. PWM or resistor, than it is too speed up a slow, quiet fan.
 
JimmiG said:
Initially the CNPS9900 did ship with a shroud, but even then Zalman recommended the removal of it. It increased both the temperature and noise level. To be honest I think the staggered orb shape is more for looks than function.


Most high-end coolers ship with the necessary components for S775 and AM2/3 mounting as well. Just keep in mind that the AMD socket is rotated 90 degrees compared to Intel so the cooler might have to face upwards instead of blowing the air toward the rear exhaust fans.
Click to expand...


True but their performance doesn't necessarily correlate to how they perform with Intels.
 
Mangudai said:
True but their performance doesn't necessarily correlate to how they perform with Intels.
Click to expand...

Is there really a difference? With the heat spreader on, Intel and AMD CPUs looks pretty much identical on the outside (except one has pads under it, the other pins). Die sizes are also similar as is thermal power. Then again, the CNPS9900 doesn't seem to be as great with AMD CPUs as it appears to be with the i7 in this review...
 
Last edited:
I also did not get the results with the 9900 (comparitivly speaking) that is shown here. I found that the new Zalman 10X was better by 6 degrees or so.
That brings up anothor point, this is a new review but the coolers tested do not include some of the newer ones.
 
Here are some shots of the Hurricane from different sides. Take note of the last two photos. Notice that on the bottom and the top (all the way through the fins) are these two extra holes that appear to have been reserved for extra heat pipes. Perhaps TTIC will release a future version of this cooler with better performance or perhaps internal testing revealed they were unnecessary. Either way it caught my attention.
Click to expand...

check the platform compatibility ;) those holes are to secure the HSF on AMD platforms ;)
 
i wish they'd give us more info on the sound aspect. some fans make high pitched noises and others make lower less annoying sounds. just posting DBs isn't enough.
Click to expand...
That's what the A specification is supposed to offset compared with normal dB's.

That most likely wouldn't be accurate though. The test setup would have to be identical. Ambient room temperature, case air flow, specific processor used in the test bed, etc would all have to be the exact same. You can be as meticulous in your testing procedures as you want but there's always going to be some factor that's a little different when you don't test them all at the same time. I doubt many hardware review sites have the capabilities to do a 30+ heatsink round up in a single day on a single test bed.
Click to expand...
They don't have to. Air conditioning solves a lot of problems.
 
undertheradar said:
Still no IFX-14 in these reviews... what, does everyone fear the king?

I find the results of the Zalman odd... almost 'please welcome our new sponsor, ________' odd.... lol. But seriously...
http://www.bit-tech.net/hardware/cooling/2009/02/19/ga-1366-cpu-cooler-group-test/7

http://www.xbitlabs.com/articles/coolers/display/core-i7-coolers-roundup_18.html

http://www.frostytech.com/articleview.cfm?articleid=2362&page=3

Something tells me that the degree of control in this experiment was not up to snuff. In most other tests, the 9900 doesnt even come close to top dog.
Click to expand...

agreed. ive learned to only stick to forums when browsing H. :)
 
how much better does the Rev B Megahalems perform compared to its first revision. i can't find any accurate or credible reviews anywhere online? does anyone know?
 
Thanks for the great reviews. Just about to start up my next build (waiting for all the parts to come in). Very handy to know before doing the CPU install.
 
You must log in or register to reply here.
Back
Top