Overclock.net banner

1 - 20 of 76 Posts

·
Registered
Joined
·
1,286 Posts
Discussion Starter · #1 ·
The test were done on my own test bench to find the differences between the heatsinks while using the same fans. I chose to do the test using the retail NF-A15 PWM which only goes up to 1200rpm instead of the 1500rpm versions that come with the NH-D15 since the 1500rpm version's minimum rpm is ~680-700rpm which is too loud for "silent" applications.

Testing methodology:
Setup is on open test bench using a 4670k delidded and OCed to 4.3ghz @ 1.24V. Sorry by CPU isn't stable no matter what vcore it's at.

Paste used for the NH-D14 was IC Diamond and NH-D15 was Coollaboratory Liquid Ulra since both heatsinks are different machined base. The D14 had a rougher base which suited IC Diamond, while D15 was almost smooth which suited Liquid Ultra. I had initially tried IC Diamond on the D15 and temps were significantly worse than the D14. DO NOT USE VISCOUS PASTE FOR THE D15!

For load testing, I ran prime95 small FFT for 6 mins on each test. Small FFT achieves high core temp in a short period of time. Temperatures were then averaged across the 4 cores then subtracted from the ambient temp.


 

·
Premium Member
Joined
·
27,617 Posts
Looks like a 3.5-4.5c difference between NH-D14 and NH-D15?
 

·
Registered
Joined
·
1,286 Posts
Discussion Starter · #3 ·
Quote:
Originally Posted by doyll View Post

Looks like a 3.5-4.5c difference between NH-D14 and NH-D15?
Yup, and that's the heatsink difference which is quite significant. I'm actually a bit disappointed Noctua did not make the heatsink larger by using more fins to fully take advantage of the fan. Probably could have dropped the temps by another 1.5C, but the babies would cry about memory clearance.
mad.gif


Worthy upgrade when you consider I only paid $70 for this D15. Good job Noctua for making an improvement, bad job for not taking advantage of it's full potential for the sake of compatibility.
 

·
Premium Member
Joined
·
9,160 Posts
Quote:
Originally Posted by ohhgourami View Post

4670k delidded and OCed to 4.3ghz @ 1.24V
Hrm, that looks decent. I wonder how further it could go on another cpu. I got hints from another review that a main obstacle might be 'jumps' of temp, e.g. the avg temperature being around 70-80 but one core reaching much higher and crashing the thing.
 

·
Registered
Joined
·
1,286 Posts
Discussion Starter · #5 ·
Quote:
Originally Posted by fateswarm View Post

Hrm, that looks decent. I wonder how further it could go on another cpu. I got hints from another review that a main obstacle might be 'jumps' of temp, e.g. the avg temperature being around 70-80 but one core reaching much higher and crashing the thing.
Maybe. The D15 has provided me the most even temps of the heatsinks I've tried. Of the 10+ mountings I've done with the D14, I've always had a delta 8C between cores. I only got a delta 4C on the D15. Maybe lapping the heatsink and chip would fix that.
 

·
Retired Reviewer
Joined
·
10,099 Posts
To compare heatsinks you really ought to use the same TIM. In this case, for example, Noctua's own TIM would probably be most appropriate. I would have use GC-Extreme, but that's me. The Noctua paste should suit both heatsinks since Noctua would have made sure thry were compatible befor releasing the heatsinks.
 

·
Registered
Joined
·
1,286 Posts
Discussion Starter · #8 ·
Quote:
Originally Posted by ehume View Post

To compare heatsinks you really ought to use the same TIM. In this case, for example, Noctua's own TIM would probably be most appropriate. I would have use GC-Extreme, but that's me. The Noctua paste should suit both heatsinks since Noctua would have made sure thry were compatible befor releasing the heatsinks.
I initially used IC Diamond for both. And IC Diamond has provided by the best temps for the D14 compared many other pastes. Too bad the D15 had horrible temps with IC Diamond. I had to use 1500rpm fans just to match the D14 using 1200rpm. I get what you mean by using the same paste, but it's just not optimal since the bases are different.

Ideally, we would lap both heatsinks and use Liquid Ultra.
 

·
Registered
Joined
·
83 Posts
I agree with ehume, it might be that the few degrees of differences in your results are caused by the superiority of liquid metal TIM over any other type of TIM and not really by the difference in the capabilities of D14 and D15.
 

·
Registered
Joined
·
1,286 Posts
Discussion Starter · #12 ·
Quote:
Originally Posted by veryrarium View Post

I agree with ehume, it might be that the few degrees of differences in your results are caused by the superiority of liquid metal TIM over any other type of TIM and not really by the difference in the capabilities of D14 and D15.
When I compared Liquid Ultra and IC Diamond on the D14, Liquid Ultra was worse by 1C.
 

·
Registered
Joined
·
657 Posts
Quote:
Originally Posted by ohhgourami View Post

When I compared Liquid Ultra and IC Diamond on the D14, Liquid Ultra was worse by 1C.
It's not a good methodology to compare heatsinks using different pastes, so the validity of your experiment is questionable.
Your experiment seems to agree with what other reviewers are saying about the D15 vs. the D14 so I'm not going to completely throw your results out the window.
 

·
Premium Member
Joined
·
27,617 Posts
Is this just one mounting and run or several mounting with several runs on each mounting?
 

·
Registered
Joined
·
1,286 Posts
Discussion Starter · #15 ·
The mounting on the D14 is from the last of 5+ mountings with IC Diamond. All 5 gave the same exact temps.

I only did one mounting job with the D15 each with IC Diamond and Liquid Ultra. I'm currently out of Liquid Ultra so I can't repeat the test. I'm skeptical the IC Diamond application was done poorly since the footprint looks good; Liquid Ultra is a mandatory spread and this cooler is very easy to seat.

I'll admit the use of different paste for the final data is questionable, but I believe these paste gave the heatsinks their best respective temps and that is what I wanted to test - the differences in temps when they are performing their best when eliminating differences in fans used.
 

·
Premium Member
Joined
·
27,617 Posts
I agree. Your results are similar to others comparing the NH-D14 and NH-D15.
Bottom line is NH-D15 brings Noctua back on par with other top coolers like PH-TC14PE, Silver Arrow IB-E, Cryorig R1 Ultimate, etc.

Considering the age of NH-D14, being only a few degrees warmer than newest releases is impressive. But than if original Silver Arrow was still being made...
wink.gif
 

·
PC Enthusiast
Joined
·
1,093 Posts
Thanks for the review + Rep.

Is the d15 worth the extra $20(approx) over the D14? or is the d14 still a good buy?
 
1 - 20 of 76 Posts
Top