Industry heatsink+TIM testing@frostytech - Page 2

BillA · 01-08-2006, 10:36 PM

"It is pretty sad though when end user reviews seem more usefull than bench testing."

a wake-up call ?

bigben2k · 01-10-2006, 02:20 PM

Shoot, I'll be happy when I can measure something, anything. I'm not concerned with the geometry of the processor heat, simply because an IHS and the baseplate of a block will (IMO) smooth things out to be near-indistinguishable.

However, there is some anecdotal evidence that a Storm type design, whose cooling is highly localized, may result in a better overclock. Is there any truth to it? Knowing that the baseplate of Storm is ~1 mm thick, and we know the IHS (~1.4mm for Intel (~1.65mm for LGA), ~2.4mm for AMD), is any localization of the cooling effect going to have any impact, or is it all smoothed out in the thickness?

I'm not sure how I would even go about testing that, unless I used various sizes of heat dies.

BillA · 01-10-2006, 02:32 PM

I do believe that w/o an IHS the Storm is better, but I too have no data
one would postulate a smaller source

Les · 01-10-2006, 04:12 PM

Quote:

Originally Posted by BillA

I do believe that w/o an IHS the Storm is better, but I too have no data
one would postulate a smaller source

"one would postulate a smaller source"
Whilst I am postulating that with progressively smaller sources the Storm gets worse(link and link).The attachment in second link is particularly illustrative.

BillA · 01-10-2006, 04:15 PM

hmmm
are we not assuming at least some spreading ?
lord this is a mess Les

Les · 01-10-2006, 04:48 PM

No spreading in my die.
All are top surface sources and Waterloo based.
Incoherent is looking at bottom die sources.
Preliminarily work does not immediately support but is early days
Monitoring progress with Waterloo
It is hard going.
Direction and extent of work very uncertain.
May be most productive to concentrate on thermal-geometries for which data is/maybe available - Fluxbloc.

It is a mess.
However, not unhappy that slotted IHS is correct for determining pecking order of heatsinks.
Whether the performance of a heatsink is critical in thermal design is another matter

BillA · 01-10-2006, 05:00 PM

yes, real data is needed to prove the model

"Whether the performance of a heatsink is critical in thermal design is another matter"
not understood, a chicken and egg thing ?
all temps change with the sink resistance

Les · 01-10-2006, 05:08 PM

Yes, all temperatues change.
How much is the important question.
Intel and AMD control the gradients below TIM2

01-08-2006, 10:36 PM	#26
BillA CoolingWorks Tech Guy Formerly "Unregistered" Join Date: Dec 2000 Location: Posts: 2,371.493,106 Posts: 4,440	Re: Industry heatsink+TIM testing@frostytech "It is pretty sad though when end user reviews seem more usefull than bench testing." a wake-up call ?

01-10-2006, 02:20 PM	#27
bigben2k Responsible for 2% of all the posts here. Join Date: May 2002 Location: Texas, U.S.A. Posts: 8,302	Re: Industry heatsink+TIM testing@frostytech Shoot, I'll be happy when I can measure something, anything. I'm not concerned with the geometry of the processor heat, simply because an IHS and the baseplate of a block will (IMO) smooth things out to be near-indistinguishable. However, there is some anecdotal evidence that a Storm type design, whose cooling is highly localized, may result in a better overclock. Is there any truth to it? Knowing that the baseplate of Storm is ~1 mm thick, and we know the IHS (~1.4mm for Intel (~1.65mm for LGA), ~2.4mm for AMD), is any localization of the cooling effect going to have any impact, or is it all smoothed out in the thickness? I'm not sure how I would even go about testing that, unless I used various sizes of heat dies. __________________ WBTA (Water Block Testing Alliance) Nordic Hardware WC 101

01-10-2006, 02:32 PM	#28
BillA CoolingWorks Tech Guy Formerly "Unregistered" Join Date: Dec 2000 Location: Posts: 2,371.493,106 Posts: 4,440	Re: Industry heatsink+TIM testing@frostytech I do believe that w/o an IHS the Storm is better, but I too have no data one would postulate a smaller source

01-10-2006, 04:15 PM	#30
BillA CoolingWorks Tech Guy Formerly "Unregistered" Join Date: Dec 2000 Location: Posts: 2,371.493,106 Posts: 4,440	Re: Industry heatsink+TIM testing@frostytech hmmm are we not assuming at least some spreading ? lord this is a mess Les

01-10-2006, 04:48 PM	#31
Les Cooling Savant Join Date: Oct 2001 Location: Wigan UK Posts: 929	Re: Industry heatsink+TIM testing@frostytech No spreading in my die. All are top surface sources and Waterloo based. Incoherent is looking at bottom die sources. Preliminarily work does not immediately support but is early days Monitoring progress with Waterloo It is hard going. Direction and extent of work very uncertain. May be most productive to concentrate on thermal-geometries for which data is/maybe available - Fluxbloc. It is a mess. However, not unhappy that slotted IHS is correct for determining pecking order of heatsinks. Whether the performance of a heatsink is critical in thermal design is another matter

01-10-2006, 05:00 PM	#32
BillA CoolingWorks Tech Guy Formerly "Unregistered" Join Date: Dec 2000 Location: Posts: 2,371.493,106 Posts: 4,440	Re: Industry heatsink+TIM testing@frostytech yes, real data is needed to prove the model "Whether the performance of a heatsink is critical in thermal design is another matter" not understood, a chicken and egg thing ? all temps change with the sink resistance

01-10-2006, 05:08 PM	#33
Les Cooling Savant Join Date: Oct 2001 Location: Wigan UK Posts: 929	Re: Industry heatsink+TIM testing@frostytech Yes, all temperatues change. How much is the important question. Intel and AMD control the gradients below TIM2

Currently Active Users Viewing This Thread: 1 (0 members and 1 guests)