How important CPU Block Fin density?

[Hexenhammer.]

Hi,

A new corsair LGA1700/1200 block has 110 fins [old one had 60], Heatkiller PRO IV has 78, Bykski/Barrow blocks [the good ones, not the bottom cheap], have 48-60

How important is the fin density inside the CPU block? Will this XC7 RGB PRO  [no reviews yet] block offer better cooling versus others?

 

Thanks

 

[Guest]

More Fins= More surface area= Better thermals.

[Mick Naughty]

Assuming the fin length and depth is comparable across the brands/models. Wouldnt put too much though into it. No need to chance 1-3c based off other peoples results. 

[HM-2]

3 hours ago, WickedThunder86 said:

More Fins= More surface area= Better thermals.

All else being equal, this. However there are a myriad of other factors which mean that it's not very good in isolation as a metric for determining performance. 

[bowrilla]

Isn't the IHS size of Alder Lake quite a lot bigger than with previous chips? So you obviously will have a higher total fin count even with the same fin density.

 

Generally speaking: more surface = more heat transfer = better.

 

But you also need to get your water inbetween those fins. The smaller the gaps the higher the restriction. The deeper the gaps the harder it is to get your water down through the gap (-> jet plates -> higher restriction).

[Hexenhammer.]

7 hours ago, bowrilla said:

Isn't the IHS size of Alder Lake quite a lot bigger than with previous chips? So you obviously will have a higher total fin count even with the same fin density.

 

Generally speaking: more surface = more heat transfer = better.

 

But you also need to get your water inbetween those fins. The smaller the gaps the higher the restriction. The deeper the gaps the harder it is to get your water down through the gap (-> jet plates -> higher restriction).

 

So should i go for the new 110 fin Corsair XC7 RGB PRO that was made for Adler lake instead of using old 78 fin Heatkiller IV Pro?

 

 

[Hexenhammer.]

7 hours ago, HM-2 said:

All else being equal, this. However there are a myriad of other factors which mean that it's not very good in isolation as a metric for determining performance. 

 

I have 2 rads, 360GTX and 480GTS and dual pumps, DDC build into the reservoir and D5 outside

[HM-2]

35 minutes ago, Hexenhammer. said:

 

I have 2 rads, 360GTX and 480GTS and dual pumps, DDC build into the reservoir and D5 outside

I meant factors between blocks, rather than within the full loop. 

[bowrilla]

1 hour ago, Hexenhammer. said:

 

So should i go for the new 110 fin Corsair XC7 RGB PRO that was made for Adler lake instead of using old 78 fin Heatkiller IV Pro?

 

 

I'd say that's the best idea since the IHS is a bit larger and you want the cold plate to cover as much surface as possible on the IHS. That's exactly why your regular AiO designed for AM4 or LGA1200 is not a good fit for Threadripper on TR4 even though there might be a suitable mounting kit.

[AnonymousGuy]

Alderlake die size isn't that large so the Heatkiller IV Pro fin area would already cover it.  Don't bother replacing blocks.  The IHS being larger is somewhat irrelevant since most of the heat transfer happens through the IHS directly in contact with the die.

 

We're at the point where really all CPU blocks that are good (which the Heatkiller still is) are more or less the same.  Having crazy high fin density just leads to a really restrictive block without much thermal benefit.  That's why a block from 2015 still hasn't been "upgraded".

[AnonymousGuy]

On 11/16/2021 at 12:47 PM, bowrilla said:

I'd say that's the best idea since the IHS is a bit larger and you want the cold plate to cover as much surface as possible on the IHS. That's exactly why your regular AiO designed for AM4 or LGA1200 is not a good fit for Threadripper on TR4 even though there might be a suitable mounting kit.

You're mixing IHS and die size in an improper way.  Threadripper used a lot of substrate surface area because of 4 discrete silicon chips.  So yes it led to a larger IHS but it was also a huge silicon surface area which is really why you needed larger fin areas to cover.  I don't think there's any single die chip that exceeds the fin area of a waterblock, mainly because it isn't economically viable to fab a piece of silicon that large with how defect density works.  (I think the 3080/3090 are probably the largest.  Maybe Xeon Phi or Itanium were similarly huge but no one here uses those.)

 

Alderlake is basically the same size as anything since Ivy Bridge.  The IHS changing dimensions is due (probably) to pinout requirements for IO like DDR5 and not because of the actual silicon getting larger:
 

 

Ivy bridge:
 

 

And meanwhile Zen 3 (which BTW doing it like this fucks over your profit margins because you're paying 2x more just for the silicon fabwork and also they're paying it to a contractor who needs to make a 20% profit vs. Intel paying it to themselves):