Recommended Heatsinks updated

Want to talk about one of the articles in SPCR? Here's the forum for you.
quest_for_silence
Posts: 5275
Joined: Wed Jun 13, 2007 10:12 am
Location: ITALY

Re: Recommended Heatsinks updated

Post by quest_for_silence » Wed Sep 17, 2014 7:53 am

MikeC wrote:Data for D14 comes from... http://www.silentpcreview.com/article1020-page6.html

Then it shouldn't apply, Mike, as the table title says "Tested on OC'd 130W TDP 1366 i7": if I'm not wrong the data you linked are referred to stock settings (no OC).

There's a minor, similar issue with the Prolimatech Genesis data, the figures should refer to the setup with 2 x Prolimatech Red Vortex 14 LED, so that they shouldn't apply as in the beginning of the article there's written: "We prefer to assess the cooling power of the heatsink with a quiet reference fan... There are exceptions with units that are designed with the specific fan as part and parcel of the package" (and the exception should not be the case of any Prolimatech heatsink, which doesn't bundle any fan).

Thanks in advance for your understanding in the matter.

MikeC
Site Admin
Posts: 12285
Joined: Sun Aug 11, 2002 3:26 pm
Location: Vancouver, BC, Canada
Contact:

Re: Recommended Heatsinks updated

Post by MikeC » Wed Sep 17, 2014 8:17 am

quest_for_silence wrote:
MikeC wrote:Data for D14 comes from... http://www.silentpcreview.com/article1020-page6.html

Then it shouldn't apply, Mike, as the table title says "Tested on OC'd 130W TDP 1366 i7": if I'm not wrong the data you linked are referred to stock settings (no OC).

There's a minor, similar issue with the Prolimatech Genesis data, the figures should refer to the setup with 2 x Prolimatech Red Vortex 14 LED, so that they shouldn't apply as in the beginning of the article there's written: "We prefer to assess the cooling power of the heatsink with a quiet reference fan... There are exceptions with units that are designed with the specific fan as part and parcel of the package" (and the exception should not br the case of any Prolimatech heatsink, which doesn't bundle any fan).

Thanks in advance for your understanding in the matter.
Actually it comes from the table w/ the header: Noctua NH-D14 w/ Stock Fans, i7-965 default settings.
I need to clean up a few other things in the piece; hopefully it can be done in the next couple hrs. One thing I do want to emphasize is that for these lists, it is the performance with STOCK fans that matters the most -- except in the case where fans are not provided. (Such cases are increasingly rare.) The assumption is -- esp. for expensive coolers -- that most buyers will use the stock fans unless they have poor acoustic quality. In most cases, when spun slow to sub-20 dB levels, it's not much of an issue even if SQ is less than pristine.

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

Re: Recommended Heatsinks updated

Post by quest_for_silence » Wed Sep 17, 2014 8:40 am

MikeC wrote:I need to clean up a few other things in the piece; hopefully it can be done in the next couple hrs.

Thanks for all your efforts, and if you ever had to think that any small help, in the form of some proofreading by an untalented amateur, might be somehow welcome, just drop a PM (wow, what a complicated phrase I wrote!).

MikeC
Site Admin
Posts: 12285
Joined: Sun Aug 11, 2002 3:26 pm
Location: Vancouver, BC, Canada
Contact:

Re: Recommended Heatsinks updated

Post by MikeC » Wed Sep 17, 2014 10:05 am

MikeC wrote:Guys, my apologies for posting this w/o more thorough review. My self-imposed deadline caused errors, obviously. I must have missed the Kotetsu, and I did forget the nofan. Data for D14 comes from... http://www.silentpcreview.com/article1020-page6.html
Actually, on second look, the Kotetsu is in the list -- just a bit lower down than you think.

Also, I'm adding a column on whether the heatsink is top-blow or side-flow. There were some immediate complaints about this.

Fanless HS are in, too.

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

Re: Recommended Heatsinks updated

Post by xan_user » Wed Sep 17, 2014 3:57 pm

MikeC wrote:
Fanless HS are in, too.
any plans at all to take a look at the zalman fx70 ? http://www.zalman.com/global/product/Pr ... hp?Idx=872

MikeC
Site Admin
Posts: 12285
Joined: Sun Aug 11, 2002 3:26 pm
Location: Vancouver, BC, Canada
Contact:

Re: Recommended Heatsinks updated

Post by MikeC » Wed Sep 17, 2014 4:08 pm

xan_user wrote:
MikeC wrote:
Fanless HS are in, too.
any plans at all to take a look at the zalman fx70 ? http://www.zalman.com/global/product/Pr ... hp?Idx=872
Actually hadn't seen this before. Looks worth a look, thanks!

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

Re: Recommended Heatsinks updated

Post by xan_user » Wed Sep 17, 2014 4:42 pm

MikeC wrote:
xan_user wrote:
MikeC wrote:
Fanless HS are in, too.
any plans at all to take a look at the zalman fx70 ? http://www.zalman.com/global/product/Pr ... hp?Idx=872
Actually hadn't seen this before. Looks worth a look, thanks!
8)
Its not often I want to buy something based just on it looks...but damn, it would look mean sticking out of a m350 or SilverStone PT13!

Abula
Posts: 3662
Joined: Tue Nov 02, 2004 12:22 pm
Location: Guatemala

Re: Recommended Heatsinks updated

Post by Abula » Wed Sep 17, 2014 4:58 pm

Although already discuss seems like the NH-D14 too good in the chart.

MikeC
Site Admin
Posts: 12285
Joined: Sun Aug 11, 2002 3:26 pm
Location: Vancouver, BC, Canada
Contact:

Re: Recommended Heatsinks updated

Post by MikeC » Wed Sep 17, 2014 5:38 pm

Abula wrote:Although already discuss seems like the NH-D14 too good in the chart.
iirc, it was reviewed during transition between the previous 1366 board and the current one... There may be consistency issues. We'd have to retest to be sure... if we had one still. But then it hardly matters, the thing's will be discontinued soon anyway now that D15 is here. :lol:

Seriously, tho, the differences are still small enough not to be that big a deal, and anyone who reads the review will see the 120mm fan (w/ 140mm fan) causes some acoustic issues -- even tho it's at a low level.

nizer
Posts: 32
Joined: Tue May 08, 2012 12:27 pm
Location: Princeton, US

Re: Recommended Heatsinks updated

Post by nizer » Wed Sep 17, 2014 10:53 pm

Thank you for the update of the recommended heatsinks section, an important part of the site!

Mrc112
Posts: 59
Joined: Mon Mar 12, 2007 6:38 am
Location: Belgium

Re: Recommended Heatsinks updated

Post by Mrc112 » Thu Sep 18, 2014 6:26 am

The nh-d15 link is pointing to the nh-d14 article ;-)

MikeC
Site Admin
Posts: 12285
Joined: Sun Aug 11, 2002 3:26 pm
Location: Vancouver, BC, Canada
Contact:

Re: Recommended Heatsinks updated

Post by MikeC » Thu Sep 18, 2014 9:00 am

Noctua D-15 link corrected.

Also, spurred by Abula & Quest-for-Silence's dogged insistence on data integrity, I examined all the testing results for the Noctua D-14 and concluded that the top-of-list results previously posted were obtained from the previous 1366 platform. This was abandoned after a few months as the board developed thermal inconsistencies. Unfortunately, coolers were retested on the new (current) board only with the reference Nexus fans, and the D14 results were not quite as strong (though results for all the retested HS seem to be slightly poorer). Hence, the D14 has been repositioned.

It's safe to say the D14 is still a top performer.

ps -- if you examine all the data points across the SPL range, you'll see how difficult it is to rank the coolers, as they often differ by just a degree up at one SPL and then down at another SPL. Stated many times: 1 degree or 1 dBA = virtually nothing.

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

Re: Recommended Heatsinks updated

Post by quest_for_silence » Thu Sep 18, 2014 9:20 am

xan_user wrote: Its not often I want to buy something based just on it looks...but damn, it would look mean sticking out of a m350 or SilverStone PT13!

From what I've seen the relevant thermal collapse level is situated around an 80W load, slightly less than the more massive FX100.
I don't expect it can run fanless as the NoFan.

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

Re: Recommended Heatsinks updated

Post by quest_for_silence » Thu Sep 18, 2014 9:22 am

MikeC wrote:Noctua D-15 link corrected.

Also, spurred by Abula & Quest-for-Silence's dogged insistence on data integrity

Thanks a lot for your commendable commitment, MikeC, and please forgive me for having bothered you that much.

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

Re: Recommended Heatsinks updated

Post by xan_user » Thu Sep 18, 2014 9:25 am

quest_for_silence wrote:
xan_user wrote: Its not often I want to buy something based just on it looks...but damn, it would look mean sticking out of a m350 or SilverStone PT13!

From what I've seen the relevant thermal collapse level is situated around an 80W load, slightly less than the more massive FX100.
I don't expect it can run fanless as the NoFan.
since ive only got a theoretical max 65watts to deal with, that should be fine. especially exposed and outside of the case, and entirely isolated from any GPU heat.

MikeC
Site Admin
Posts: 12285
Joined: Sun Aug 11, 2002 3:26 pm
Location: Vancouver, BC, Canada
Contact:

Re: Recommended Heatsinks updated

Post by MikeC » Thu Sep 18, 2014 9:28 am

quest_for_silence wrote:Thanks a lot for your commendable commitment, MikeC, and please forgive me for having bothered you that much.
No bother, you've helped to improve my work!

Olle P
Posts: 711
Joined: Tue Nov 04, 2008 6:03 am
Location: Sweden

Re: Recommended Heatsinks updated

Post by Olle P » Thu Sep 18, 2014 11:12 pm

This sentence makes me a bit puzzled:
The term "cooler" (which really irks my language logic) is actually apropos here, as both "air only" heatsinks and all-in-one water cooling devices are in the mix.
Strictly speaking today just about all coolers use water, mostly by heat pipes!
  1. "Air only" to me only apply to my first computer, some 20 years ago. That one didn't have a heatsink on the CPU (a 66MHz IBM SLC2, IIRC) so the air had free access to the CPU's capsule.
  2. Next step in the development of coolers was to apply a "single piece of metal" (typically aluminium) heatsink. (Eventually upgraded by mounting a fan on it.)
  3. Next design development was the "double metal" where copper is used to transfer the heat from the CPU to the aluminium heatsink.
  4. After that (forced) water cooling became popular. Here pump driven water absorb heat at the hot spots and dump it to cooling fins in a "radiator" (that is actually a "convector" since the vast majority of the heat isn't given off as black body radiation.)
  5. Nowadays it's almost 100% thermodynamically driven water cooling (heat pipe) that transfer heat from the hot spot to the cooling fins.

nagi
Posts: 125
Joined: Tue Aug 31, 2010 6:48 pm
Location: Outside the box

Re: Recommended Heatsinks updated

Post by nagi » Fri Sep 19, 2014 1:15 am

Nice!


On the future VGA cooler update: I have the Accelero Xtreme IV, and there is one big problem with it: VRM cooling. Many users go ahead and cut the original heatsink in half, and put it under the Xtreme IV, and then cooling is absolutely a beast. Seriously, in most games you won't hear the fans ramp up to any noticeable speed. (Note that it also doesn't offer the same cooling to the memory chips, but there is no real solution to that except for installing your own small sinks onto them.)

MikeC
Site Admin
Posts: 12285
Joined: Sun Aug 11, 2002 3:26 pm
Location: Vancouver, BC, Canada
Contact:

Re: Recommended Heatsinks updated

Post by MikeC » Fri Sep 19, 2014 6:26 am

Olle P wrote:This sentence makes me a bit puzzled:
The term "cooler" (which really irks my language logic) is actually apropos here, as both "air only" heatsinks and all-in-one water cooling devices are in the mix.
Strictly speaking today just about all coolers use water, mostly by heat pipes!
I should have added -- "I much prefer heatsink", which is more precise, imo. Cooler has so many definitions; I think of it as the chest for beer & sandwiches, the opposite of warmer, a drink or jail. But heatsink, unlike cooler, doesn't include watercooling devices. Hence my comment that cooler is apropos.

OK, my language logic isn't quite complete. From http://dictionary.reference.com/browse/cooler:

cooler:-
1. a container or apparatus, as an insulated chest, in which something may be cooled or kept cool.
2. anything that cools or makes cool; refrigerant.
3. an air conditioner.
4. a tall drink, consisting of liquor, soda, and a fruit garnish.
5. water cooler.
6. the cooler, Slang. jail: "He was in the cooler for three months for petty theft."
7. Ice Hockey Slang. penalty box.

I still don't have to like it though. It's just... sloppy. :lol: :roll:

Das_Saunamies
*Lifetime Patron*
Posts: 2000
Joined: Tue May 15, 2007 1:39 am
Location: Finland

Re: Recommended Heatsinks updated

Post by Das_Saunamies » Sat Sep 20, 2014 12:09 am

Nice to see a fresh update, the Recommended articles could certainly use frequent refreshes.

lb_felipe
Posts: 270
Joined: Wed Dec 24, 2008 12:21 pm
Location: São Paulo, SP, Brasil

Re: Recommended Heatsinks updated

Post by lb_felipe » Sun Sep 28, 2014 11:19 pm

Where is Megahalems?

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

Re: Recommended Heatsinks updated

Post by quest_for_silence » Mon Sep 29, 2014 12:32 am

lb_felipe wrote:Where is Megahalems?

http://www.silentpcreview.com/Recommended_Heatsinks (it is ranked 10th place)

MikeC
Site Admin
Posts: 12285
Joined: Sun Aug 11, 2002 3:26 pm
Location: Vancouver, BC, Canada
Contact:

Re: Recommended Heatsinks updated

Post by MikeC » Mon Sep 29, 2014 1:24 am

quest_for_silence wrote:
lb_felipe wrote:Where is Megahalems?

http://www.silentpcreview.com/Recommended_Heatsinks (it is ranked 10th place)
TBH, it wasn't there before, I'd missed it before lb_felipe pointed it out.

I know you already know this, Luca, but just to make it perfectly clear... The idea that it's in 10th place isn't quite accurate -- the table isn't entirely linear because of the many fan speed/noise test points. Look up and down 5 places or more and there's hardly any change at some of those test points. My take on the list is that if you have a CPU rated 130W TDP or less, and you want 17 dBA or lower SPL, chances are, any of the top 10 or so air coolers will serve you fine. It's not until you listen to the sound quality of the fan that you could nitpick among them, you can't differentiate just by SPL, they are just too close. And even there, if the SPL is low enough, the SQ really won't matter anyway.

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

Re: Recommended Heatsinks updated

Post by quest_for_silence » Mon Sep 29, 2014 1:58 am

MikeC wrote:The idea that it's in 10th place isn't quite accurate

Just to be more clear, mine was just a topological information.

Broadly speaking, as for any PC part, figures actually don't explain what really matters.

Olle P
Posts: 711
Joined: Tue Nov 04, 2008 6:03 am
Location: Sweden

Re: Recommended Heatsinks updated

Post by Olle P » Wed Oct 01, 2014 12:27 am

MikeC wrote:I should have added -- "I much prefer heatsink", which is more precise, imo. Cooler has so many definitions; ...
I think you missed my point.
It was this part of the sentence: "... 'air only' heatsinks and all-in-one water cooling devices...", that triggered me.
* There's no "air only" (no heatsink!) in practical use with CPUs.
* Most CPU cooling devices nowadays use water cooling, typically by means of heat pipe.

Das_Saunamies
*Lifetime Patron*
Posts: 2000
Joined: Tue May 15, 2007 1:39 am
Location: Finland

Re: Recommended Heatsinks updated

Post by Das_Saunamies » Wed Oct 01, 2014 7:58 am

Olle P wrote:
MikeC wrote:I should have added -- "I much prefer heatsink", which is more precise, imo. Cooler has so many definitions; ...
I think you missed my point.
It was this part of the sentence: "... 'air only' heatsinks and all-in-one water cooling devices...", that triggered me.
* There's no "air only" (no heatsink!) in practical use with CPUs.
* Most CPU cooling devices nowadays use water cooling, typically by means of heat pipe.
I love accurate terminology, since it plays a big part in my line of work, but this here is Advanced Nitpicking; it goes so far beyond the scope of its purpose that it hinders or eliminates practical application. :lol: It reminds me of:

Image

I don't think the unobservable, miniscule amount of liquid used constitutes a defining or essential attribute of the cooling device as a whole, so it is moot from a terminological standpoint. A valid defining attribute could be the clearly observable (forced) airflow going through an 'air-cooled heatsink' or 'air cooler' (airconditioning?), same with a 'water-cooling system', arguably a complex system and not a single device per se, even in all-in-one form, or 'water cooler' (water cooler chats?) having a clearly observable circulation of coolant/water. And let's not forget the common use of cooler as something that cools food and drink, as Mike pointed out, and non-native English-speaking PC hobbyists may be keen to forget.

As much as it may jar Mike (and myself, when I run across the term again), cooler has taken on yet another meaning that is here to stay. Generic and generously applied terms, the bane and boon of texts that have to work in multiple registries (e.g. professional, common, hobbyist etc.).

doyll
Posts: 45
Joined: Wed Jul 22, 2015 3:17 am

Re: Recommended Heatsinks updated

Post by doyll » Wed Dec 16, 2015 3:59 am

Just out of curiosity,
I was wondering about using a top tier heatsink like Silver Arrow, PH-TC14PE, NH-D14, HR-22, etc to passively cool a 130w CPU. I have seen your testing of CR-95 against HR-22 and a couple other dedicated "passive" coolers. I do wonder about only CR-95 not failing. What orientation were the test carried out in? Was this only on your bench test station or were they turned on edge to simulate tower use with corresponding convection airflow between fins?

My curiosity about all of this is after using an i7 920 under PH-TC14PE in a Define R2 case modified with no back or bottom vent grills and 140mm front intakes using SilverStone air filters on front and bottom intakes, top vents open and all fans unplugged staying below 70c in a 21c room while encoding with CPU at 100% on all cores. No load on small GPU (HD 6570 GDDR5 512MB reference model w/ Scythe Setsugen cooler); not a lot of restriction to airflow and no heat added. Case is raised on caster base so has less resistance to airflow going to bottom vents and front door was open.

Sorry if this is the wrong place for this. Feel free to move or delete if it is.

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

Re: Recommended Heatsinks updated

Post by quest_for_silence » Wed Dec 16, 2015 11:08 am

doyll wrote:were they turned on edge to simulate tower use with corresponding convection airflow between fins?

As far as I know there's no simple convection in a tower enclosures equipped with some fans.

doyll wrote:staying below 70c in a 21c room while encoding with CPU at 100% on all cores.

And what about a Linpack/Prime test? How much did your encoding task last?
Broadly speaking, as far as I know it's more difficult to cool 130W of heat with an LGA115X CPU than with a LGA1366 one, due the smaller die surface.

doyll
Posts: 45
Joined: Wed Jul 22, 2015 3:17 am

Re: Recommended Heatsinks updated

Post by doyll » Wed Dec 16, 2015 11:54 am

quest_for_silence wrote: As far as I know there's no simple convection in a tower enclosures equipped with some fans.

Simple convection airflow between fins does occur if they are oriented vertically .. unless the fins are so close together air cannot flow through them. Few if any air cooling heatsinks have fins that tightly packed.
quest_for_silence wrote: And what about a Linpack/Prime test? How much did your encoding task last?
Broadly speaking, as far as I know it's more difficult to cool 130W of heat with an LGA115X CPU than with a LGA1366 one, due the smaller die surface.
I did not run Linpack/Prime, but I do know what 100% CPU load on all cores & threads is. I have compared Prime and encoding loads and resulting temps and found them to be the same .. at least so close that monitoring the actual temperature of air going into cooler and CPU they give the same delta temp.

But this is a side-track of what I was trying to find out. If you would like to discuss this further, please start a thread and let me know.

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

Re: Recommended Heatsinks updated

Post by quest_for_silence » Wed Dec 16, 2015 12:31 pm

doyll wrote:Simple convection airflow between fins does occur if they are oriented vertically

You can't assume that when any forced airflow is present, because any stack effect at that scale, with those dimensions is extremely small.

doyll wrote:I did not run Linpack/Prime

And I already understood so: nonetheless that was the failing condition in SPCR test, so for comparison purposes you should perform the same test.

doyll wrote:But this is a side-track

Though I beg to differ: you can't assume that, unless the cooler performed differently in the very same test.

Post Reply