Please recommend a cooler

Cooling Processors quietly

Moderators: NeilBlanchard, Ralf Hutter, sthayashi, Lawrence Lee

Post Reply
whispercat
Posts: 376
Joined: Tue Jul 22, 2008 6:05 pm
Location: Calgary, AB, Canada

Please recommend a cooler

Post by whispercat » Thu Jan 08, 2015 4:18 pm

Starting to research parts for a new light workstation build. My knowledge base weakness is coolers/heatsinks, I don't know too much about them and which ones work with or fit with which mobos, CPU's and cases.

Motherboard: ASRock X99 Extreme4: http://www.asrock.com/mb/Intel/X99%20Ex ... x.asp?cat=

CPU: Intel Xeon E5 1630v.3 3.7 GHz quad core (140w) http://ark.intel.com/products/family/78 ... amily#@All

Case: Fractal Define R5: http://www.fractal-design.com/home/prod ... 5-titanium

What would be the most appropriate, coolest, quietest, best fitting cooler/heatsink (active) for this configuration?

(budget $150.)

Thanks.

quest_for_silence
Posts: 5275
Joined: Wed Jun 13, 2007 10:12 am
Location: ITALY

Re: Please recommend a cooler

Post by quest_for_silence » Fri Jan 09, 2015 12:43 am

whispercat wrote:What would be the most appropriate, coolest, quietest, best fitting cooler/heatsink (active) for this configuration?

IMO any of the about first 18 recommended ones with LGA2011 fittings.

CA_Steve
Moderator
Posts: 7650
Joined: Thu Oct 06, 2005 4:36 am
Location: St. Louis, MO

Re: Please recommend a cooler

Post by CA_Steve » Fri Jan 09, 2015 11:42 am

Scythe Kotetsu: $45, top cooler with a single fan.

whispercat
Posts: 376
Joined: Tue Jul 22, 2008 6:05 pm
Location: Calgary, AB, Canada

Re: Please recommend a cooler

Post by whispercat » Fri Jan 09, 2015 1:48 pm

CA_Steve wrote:Scythe Kotetsu: $45, top cooler with a single fan.
It's extinct. Not available anywhere.

I was thinking of either the Noctua NH-U9Dx i4 (125mm) or NH-D9DX i4 3U (110mm). There are other models slightly bigger as well (e.g. NH-U12DX i4 158mm).

Question: does taller necessarily mean cooler? (given same make and model)

cerbie
Posts: 46
Joined: Sat Dec 28, 2013 3:21 pm

Re: Please recommend a cooler

Post by cerbie » Fri Jan 09, 2015 2:08 pm

whispercat wrote: Question: does taller necessarily mean cooler? (given same make and model)
They aren't the same make and model :).

Bigger means more surface area, and needing lower RPMs.

In a full sized case, I see no reason to not get a cooler made for 120mm or larger fans.

CA_Steve
Moderator
Posts: 7650
Joined: Thu Oct 06, 2005 4:36 am
Location: St. Louis, MO

Re: Please recommend a cooler

Post by CA_Steve » Fri Jan 09, 2015 2:13 pm

whispercat wrote:
CA_Steve wrote:Scythe Kotetsu: $45, top cooler with a single fan.
It's extinct. Not available anywhere.
Not extinct. On backorder at NCIX. You could ask them when they expect a restock. Or, go with the Mugen 4.
whispercat wrote:I was thinking of either the Noctua NH-U9Dx i4 (125mm) or NH-D9DX i4 3U (110mm). There are other models slightly bigger as well (e.g. NH-U12DX i4 158mm).

Question: does taller necessarily mean cooler? (given same make and model)
It comes down to:
- how well does the base mate to the CPU IHS?
- how well does the heat transfer to the fins?
- how well does the fan remove the heat from the fins?

A 150mm class tower with a great base can easily outperform a 170mm monster. That said, I wouldn't cool a 140W TDP CPU with a 92mm fanned cooler. Those two were fine with an 84-95W CPU...but I wouldn't go further.

quest_for_silence
Posts: 5275
Joined: Wed Jun 13, 2007 10:12 am
Location: ITALY

Re: Please recommend a cooler

Post by quest_for_silence » Fri Jan 09, 2015 2:20 pm

whispercat wrote:
CA_Steve wrote:Scythe Kotetsu: $45, top cooler with a single fan.
It's extinct. Not available anywhere.
quest_for_silence wrote:
whispercat wrote:What would be the most appropriate, coolest, quietest, best fitting cooler/heatsink (active) for this configuration?

IMO any of the about first 18 recommended ones with LGA2011 fittings.

whispercat
Posts: 376
Joined: Tue Jul 22, 2008 6:05 pm
Location: Calgary, AB, Canada

Re: Please recommend a cooler

Post by whispercat » Sat Jan 10, 2015 10:43 am

CA_Steve wrote: Not extinct. On backorder at NCIX. You could ask them when they expect a restock.
No ETA, and their warehouse has no orders for it. Basically extinct. I did see one on eBay though [edit: that's out of stock too]
CA_Steve wrote:Or, go with the Mugen 4.
Yes, the Mugen looks like the runner up to the Kotetsu.
CA_Steve wrote:A 150mm class tower with a great base can easily outperform a 170mm monster. That said, I wouldn't cool a 140W TDP CPU with a 92mm fanned cooler. Those two were fine with an 84-95W CPU...but I wouldn't go further.
Thanks. Will definitely look for at least a 120mm fan cooler then.

whispercat
Posts: 376
Joined: Tue Jul 22, 2008 6:05 pm
Location: Calgary, AB, Canada

Re: Please recommend a cooler

Post by whispercat » Fri Feb 06, 2015 8:48 pm

Looks like the Scythe Kotetsu has come back in stock in a few places. Yaay!

I've been pouring over the heatsink recommendation/reference list.

http://www.silentpcreview.com/Recommended_Heatsinks

I noticed for many of the coolers that there are no ambient temp rise values for certain decibel levels. I've read through the article but I still don't quite understand why, or what this means, or if one can still interpret some sort of temp numbers for those decibel levels.

Could someone explain the missing data points on the list?

quest_for_silence
Posts: 5275
Joined: Wed Jun 13, 2007 10:12 am
Location: ITALY

Re: Please recommend a cooler

Post by quest_for_silence » Sat Feb 07, 2015 5:45 am

whispercat wrote:Could someone explain the missing data points on the list?

In that table the figures shown are the recorded values from the original reviews.

For instance, with reference to the Kraken X61, you will find both the 34°C rise for the "20~19" dB column (from 19dB to 20dB), and the 37°C rise for the "18~17" dB column (from 17dB to 18dB), and nothing else because the Kraken X61 didn't recorded lower SPL figures, I mean in the "16-15" dB range (from 15dB to 16dB) and below (as you can check onto http://www.silentpcreview.com/article1408-page6.html).

whispercat
Posts: 376
Joined: Tue Jul 22, 2008 6:05 pm
Location: Calgary, AB, Canada

Re: Please recommend a cooler

Post by whispercat » Sat Feb 07, 2015 2:54 pm

So any blank spaces are just the same values as the ones next to them?

Then the Kotetsu would be the same 38° rise @ 20~19 dBA and @18~17 dBA. Is this correct?

But I notice the NZXT Havik has the same 41° entered in both 16~15 dBA and 14~13 dBA. That's why I got confused.

Image

CA_Steve
Moderator
Posts: 7650
Joined: Thu Oct 06, 2005 4:36 am
Location: St. Louis, MO

Re: Please recommend a cooler

Post by CA_Steve » Sat Feb 07, 2015 4:59 pm

So any blank spaces are just the same values as the ones next to them?
No.

They are somewhere between the value to the left and the right. If you plug the numbers into a spreadsheet and graph them, you can get a piecewise linear approximation of the expected number.

It's impossibly time consuming to set up a test where you can measure the temperature at every dB(A). So, instead, you chose a discrete number of data points. For example, what's the SPL and temp at the lowest fan rpm? What's the SPL and temp and the highest RPM? Now, choose a handful of rpms in between and measure the SPL and temp there. There's a methodology article somewhere. Go take a look.

quest_for_silence
Posts: 5275
Joined: Wed Jun 13, 2007 10:12 am
Location: ITALY

Re: Please recommend a cooler

Post by quest_for_silence » Sat Feb 07, 2015 10:18 pm

whispercat wrote:So any blank spaces are just the same values as the ones next to them?

No, they just miss.

whispercat wrote:Then the Kotetsu would be the same 38° rise @ 20~19 dBA and @18~17 dBA. Is this correct?

Absolutely no: it's just that there are not recorded values with reference to those noise ranges, the next step in noise level above 16dB is 22-23dB (so outside of the recommended chart/list): http://www.silentpcreview.com/article1391-page5.html

Anyway, you have just to read the linked original reviews, in order to try to solve your doubts.

whispercat wrote:But I notice the NZXT Havik has the same 41° entered in both 16~15 dBA and 14~13 dBA. That's why I got confused.

I can't help about this specific occurrence: in my opinion, that may be an error done while compiling the chart/list, as those figures do not seem correct to me.

whispercat
Posts: 376
Joined: Tue Jul 22, 2008 6:05 pm
Location: Calgary, AB, Canada

Re: Please recommend a cooler

Post by whispercat » Sun Feb 08, 2015 3:23 am

This was the only relevant information I could find on the methodology:
8. What about the missing data points? All coolers are tested at full fan speed as well as at 1100, 900, 700 and 550 RPM (whenever possible). This usually means that the cooling performance at maximum speed is not shown on this table, because most fans on most heatsinks run louder than 20 dBA@1m at full speed. The missing data points exist because the other tested fan RPMs did not fall into those SPL slots. Once you hone in on a particular cooler or two, it's best to study the original review for full details.


So, the missing data points would be a fraction number above or below the whole number next to it?

e.g. for the Kotetsu, @20~19 dBA the temp rise would be somewhere between 38° and 39° ?

quest_for_silence
Posts: 5275
Joined: Wed Jun 13, 2007 10:12 am
Location: ITALY

Re: Please recommend a cooler

Post by quest_for_silence » Sun Feb 08, 2015 1:03 pm

whispercat wrote:So, the missing data points would be a fraction number above or below the whole number next to it?

AFAIK no.

whispercat
Posts: 376
Joined: Tue Jul 22, 2008 6:05 pm
Location: Calgary, AB, Canada

Re: Please recommend a cooler

Post by whispercat » Sun Feb 08, 2015 7:43 pm

quest_for_silence wrote:
whispercat wrote:So, the missing data points would be a fraction number above or below the whole number next to it?

AFAIK no.
But that's what CA_Steve and the methodology seem to be saying.

CA_Steve
Moderator
Posts: 7650
Joined: Thu Oct 06, 2005 4:36 am
Location: St. Louis, MO

Re: Please recommend a cooler

Post by CA_Steve » Sun Feb 08, 2015 9:51 pm

What I'm saying is you can interpolate BETWEEN the given data points to estimate the noise level and temp. You can't create something OUTSIDE of the dataset. In the Scythe Kotetsu's case:

There's a 38C rise at 16dB(A) and a 41C rise at 14dB(A). So, it's safe to say the temp rise at 15dB(A) would be around 39.5C as a first approximation.

whispercat
Posts: 376
Joined: Tue Jul 22, 2008 6:05 pm
Location: Calgary, AB, Canada

Re: Please recommend a cooler

Post by whispercat » Mon Feb 09, 2015 2:01 pm

What would the temp rise at 20dB(A) and 18dB(A) be then? That's what I'm trying to figure out.

The Noctua NH-D15 has those data points, but not the Kotetsu. I'd like to compare them at those SPL levels.

Vicotnik
*Lifetime Patron*
Posts: 1831
Joined: Thu Feb 13, 2003 6:53 am
Location: Sweden

Re: Please recommend a cooler

Post by Vicotnik » Mon Feb 09, 2015 2:14 pm

whispercat wrote:What would the temp rise at 20dB(A) and 18dB(A) be then? That's what I'm trying to figure out.
Well, since no one has tested that, there is no way to know for sure. But since we got 38C rise at 15-16 dBA you can assume at most 38C, probably a little bit better.
whispercat wrote:The Noctua NH-D15 has those data points, but not the Kotetsu. I'd like to compare them at those SPL levels.
Why not compare this way? Kotetsu and 38C rise: 15-16 dBA. NH-D15 and 38C rise: 19-20 dBA. Winner: Kotetsu.

CA_Steve
Moderator
Posts: 7650
Joined: Thu Oct 06, 2005 4:36 am
Location: St. Louis, MO

Re: Please recommend a cooler

Post by CA_Steve » Mon Feb 09, 2015 5:28 pm

Actually - if you look on the table for the Dark Rock Pro 3 (etc) review, there's a Kotetsu datapoint at 22dB(A) -> 35C rise. So, figure somewhere between 35 and 38C rise. Call it 36C rise @ 20dB and be done :D

whispercat
Posts: 376
Joined: Tue Jul 22, 2008 6:05 pm
Location: Calgary, AB, Canada

Re: Please recommend a cooler

Post by whispercat » Mon Feb 09, 2015 9:30 pm

CA_Steve wrote:Actually - if you look on the table for the Dark Rock Pro 3 (etc) review, there's a Kotetsu datapoint at 22dB(A) -> 35C rise. So, figure somewhere between 35 and 38C rise. Call it 36C rise @ 20dB and be done :D
That's bizarre. I thought all the charts topped off at 20 dB(A). Why wasn't this info included in the Kotetsu or the recommended chart? This is so confusing.

Anyway, thanks for pointing that out. I was leaning toward the Kotetsu anyway, with the Noctua NH-D15 second.

Vicotnik
*Lifetime Patron*
Posts: 1831
Joined: Thu Feb 13, 2003 6:53 am
Location: Sweden

Re: Please recommend a cooler

Post by Vicotnik » Mon Feb 09, 2015 11:53 pm

whispercat wrote:That's bizarre. I thought all the charts topped off at 20 dB(A). Why wasn't this info included in the Kotetsu or the recommended chart? This is so confusing.
Then the chart would have to be bigger, with even more empty cells. More confusion! ;)

Look, since SPCR have such a nice uniform testing methodology it's possible to collect data from a lot of tests and present the data in a table. A glance at the table tells a lot of the results, and comparing one cooler to another based on performance is easy. If you find the chart confusing just ignore it and check the individual tests.

whispercat
Posts: 376
Joined: Tue Jul 22, 2008 6:05 pm
Location: Calgary, AB, Canada

Re: Please recommend a cooler

Post by whispercat » Tue Feb 10, 2015 10:50 am

Vicotnik wrote:
whispercat wrote:That's bizarre. I thought all the charts topped off at 20 dB(A). Why wasn't this info included in the Kotetsu or the recommended chart? This is so confusing.
Then the chart would have to be bigger, with even more empty cells. More confusion! ;)

Look, since SPCR have such a nice uniform testing methodology it's possible to collect data from a lot of tests and present the data in a table. A glance at the table tells a lot of the results, and comparing one cooler to another based on performance is easy. If you find the chart confusing just ignore it and check the individual tests.
Yeah, because having a standardized methodology with missing data points on one table for one cooler, and then including some of those points on another table for a completely different cooler, is so clear. It was so obvious that to find missing information about the Kotetsu, I should have looked up the individual review for the Rock Dark Pro 3. I mean, what was I thinking?

The chart wouldn't have to be bigger, and even it it were, it would not make understanding the missing data points any more or less confusing.

And no, most of the individual tests do not include more information on the tables for their own reviews. The Kotetsu review did not, and neither did the Nactua NH-D15.

CA_Steve
Moderator
Posts: 7650
Joined: Thu Oct 06, 2005 4:36 am
Location: St. Louis, MO

Re: Please recommend a cooler

Post by CA_Steve » Tue Feb 10, 2015 11:49 am

I may have pointed to the most recent review to show expanded results, but all of the Kotetsu results are listed in it's review. You just need to look under Test Results on page 5..

I think we may have beaten this topic to death.

Vicotnik
*Lifetime Patron*
Posts: 1831
Joined: Thu Feb 13, 2003 6:53 am
Location: Sweden

Re: Please recommend a cooler

Post by Vicotnik » Tue Feb 10, 2015 12:07 pm

whispercat wrote:Yeah, because having a standardized methodology with missing data points on one table for one cooler, and then including some of those points on another table for a completely different cooler, is so clear. It was so obvious that to find missing information about the Kotetsu, I should have looked up the individual review for the Rock Dark Pro 3. I mean, what was I thinking?
Are you saying that there is information about the Kotetsu in the Rock Dark Pro 3 review not present in the Kotetsu review?

What are you complaining about really? That the table of recommended big coolers don't contain enough data? It's fine if you ask me.

You seem not to get that the empty spaces in the table are empty. There is no data there. The coolers are tested with the stock fan and also sometimes with a reference fan, at different voltages. Noise and temp is measured at those voltages.

The temp measurements are not done at a given set of noise levels, they are done at a given set of voltage levels.

quest_for_silence
Posts: 5275
Joined: Wed Jun 13, 2007 10:12 am
Location: ITALY

Re: Please recommend a cooler

Post by quest_for_silence » Tue Feb 10, 2015 1:54 pm

Vicotnik wrote:The coolers are tested with the stock fan and also sometimes with a reference fan, at different voltages. Noise and temp is measured at those voltages.

The temp measurements are not done at a given set of noise levels, they are done at a given set of voltage levels.
Now the horse is buried, IMHO: I cannot imagine a clearer repetition/rephrasing of the FAQ #8.

whispercat
Posts: 376
Joined: Tue Jul 22, 2008 6:05 pm
Location: Calgary, AB, Canada

Re: Please recommend a cooler

Post by whispercat » Wed Feb 11, 2015 11:13 am

Vicotnik wrote: Are you saying that there is information about the Kotetsu in the Rock Dark Pro 3 review not present in the Kotetsu review?
Well, after having to look at 3 or 4 different tables now, there appears not to be.

And that's the problem. There are two or three different tables for each individual review, plus the overall recommended/reference comparison table.

For the Kotetsu, there is 1). the standard test results 2). the °C rise with stock fan 3). and the °C rise with reference fan table. And the overall recommended/reference chart.
Vicotnik wrote:What are you complaining about really? That the table of recommended big coolers don't contain enough data? It's fine if you ask me.
I am complaining about the fact that the tables don't contain enough data, the missing data isn't explained in a straightforward way, and that different data is often found on separate tables so that one has to look at three or four different tables across multiple site pages to find all the data. That's fine with you? wow.

Vicotnik wrote:You seem not to get that the empty spaces in the table are empty. There is no data there. The coolers are tested with the stock fan and also sometimes with a reference fan, at different voltages. Noise and temp is measured at those voltages.
I get that. Thanks.
Vicotnik wrote:The temp measurements are not done at a given set of noise levels, they are done at a given set of voltage levels.
I know that too. Thanks. I think it would just help if the individual comparison tables, or maybe just the overall comparison recommended/reference table would at least fill in the missing data with a placeholder value e.g. ~34-36°C instead of being left blank.

Vicotnik
*Lifetime Patron*
Posts: 1831
Joined: Thu Feb 13, 2003 6:53 am
Location: Sweden

Re: Please recommend a cooler

Post by Vicotnik » Wed Feb 11, 2015 3:11 pm

whispercat wrote:I am complaining about the fact that the tables don't contain enough data, the missing data isn't explained in a straightforward way, and that different data is often found on separate tables so that one has to look at three or four different tables across multiple site pages to find all the data. That's fine with you? wow.
All the findings are presented in the actual review, if relevant some may also be presented in the review of other coolers. A subset of the findings is presented in the recommended lists.
whispercat wrote:I think it would just help if the individual comparison tables, or maybe just the overall comparison recommended/reference table would at least fill in the missing data with a placeholder value e.g. ~34-36°C instead of being left blank.
That would be very unscientific. You cannot present your findings mixed with guesses.

xan_user
*Lifetime Patron*
Posts: 2269
Joined: Sun May 21, 2006 9:09 am
Location: Northern California.

Re: Please recommend a cooler

Post by xan_user » Wed Feb 11, 2015 4:33 pm

Vicotnik wrote:
whispercat wrote:I am complaining about the fact that the tables don't contain enough data, the missing data isn't explained in a straightforward way, and that different data is often found on separate tables so that one has to look at three or four different tables across multiple site pages to find all the data. That's fine with you? wow.
All the findings are presented in the actual review, if relevant some may also be presented in the review of other coolers. A subset of the findings is presented in the recommended lists.
whispercat wrote:I think it would just help if the individual comparison tables, or maybe just the overall comparison recommended/reference table would at least fill in the missing data with a placeholder value e.g. ~34-36°C instead of being left blank.
That would be very unscientific. You cannot present your findings mixed with guesses.
@wispercat; you could always print the charts out and pencil in (or just MSpaint in) the average values you want filled in.

Post Reply