The ultimate heatsink without thermalcompound?

[Confirmation]

if INTEL and AMD are so epic brillant smart people and they put a kind of thermal compound between the CPU and the IHS

why they dont do the IHS a " full heatsink "...

if they are so smart and full of money and 20 years doing these business, why they never tryied to do this?

full cooper heatsink, silver IHS, and some schrauben between IHS and CPU to dismount it, then with this we will just change 1 thermal compound only, and not the cpu ihs thermalcompound and late the heatsink thermal compound

seriusly, sounds like intel planned to have a separated heatsink to start a kind of deals in the market

anyone know if INTEL AND AMD made any CPU with a full IHS integrated in this last 20 years? at least they even tryied one time to do that?

 

[ao_ika_red]

anyone know if INTEL AND AMD made any CPU with a full IHS integrated in this last 20 years? at least they even tryied one time to do that?

Intel Pentium MMX 166. I still have this geezer somewhere in my garage.

 

[Confirmation]

Intel Pentium MMX 166. I still have this geezer somewhere in my garage.

ill go back to mmx 166 then, im tired of open the ihs, and put two times the thermalcompund

intel sometimes put wrong the thermalcompund "inside" the ihs and cpu this makes

core 1 temp 90
core 2 temp 90
core 3 temp 90
core 4 temp 55

just a small example

to solve this " big issue "

people has to delid the IHS and change the thermalcompund

is supposed the thermalcompund inside the cpu and IHS is designed to last much longer than other thermal compound

this means... intel some years ago, gived to you, a super special thermal compund inside the ihs and cpu, and in the heatsink a " weak thermalcompund " what a nice $$$ move from them...

same amd ofc

then, me, as user, of amd and intel, because i have both... im tired of change each 1 year ihs and heatsink thermalcompund

and believe me, a lot of people around the world dont know their thermal issues are more, for the ihs cpu thermalcompund, than the external thermalcompund..

3 things are important to have low thermal compunds

1 Silicon = if your silicon is awesome, you will use less voltage in the cpu
2 Material = nono, gold is not the best, SILVER, is the best material, but what is the problem with the silver? it has a lifetime... then GOLD dont have that problem, due to this, you can see gold in the CPU, and due to this " intel and amd thinks they have to separate cpu from heatsink?....

we are not going to live 200 years or 300 years.. to see how our silver and cooper degradates

10 years elapsed and my silver in the motherboard, cooper and aluminium in the heatsink are fine

the only way a heatsink cooper can degrade faster is for example, when you dont use a computer for 1 year, and thermalcompund is stucked there, you will see the cooper gets some blackpoints because the thermal compund does that

3 thermalcompund = as i said, intel and i didnt test yet, uses a nice thermalcompund inside the ihs and cpu

4 heatsink = have you ever tryied to just put another fan in the stock coolers? good.. you will solve a lot the high temps too.. believe or not, 2 120 fan in a intel cooler stock, makes a better air flow than the basic one

 

[mindless1]

You could randomly pick any product, in computing or not, and state there is some idealized way they could make it to where it performs better or lasts longer, but there is the cost to do so, usually some trade off, ultimately the question becomes Is it fit for its purpose?

Many times (far too many to count) I have looked at some product failure and thought to myself, if only they had spent 20 cents more on a better spring, or $1 more on a better connector, or whatever the earliest failure point was, could have been delayed with better design and yet if you multiply that level of design times the # of components in some complex products, they would soon become much more expensive, especially very complex things like automobiles.

I have always felt that some AMD and Intel heatsinks were marginal performers, but they do their job and an enthusiast knows you can buy aftermarket 'sinks. I have not only had no problem keeping CPUs cool with aftermarket 'sinks, I've even been able to overvolt and overclock them which is going beyond Fit For Purpose.

The vast majority of Intel's market has no problem with how the IHS is involved so there is no reason for them to change it. They do not want the price of CPUs to rise in order to meet your idealized construction methods.

Are you old enough that you recall that for a period, both AMD and Intel had open die CPUs without a IHS? These CPUs were more prone to damage from a ham-fisted heatsink installer, it was more important that the bottom of the heatsink be near perfectly flat (most were not at the time and cheap heatsinks today still aren't), and many consider the move back to an IHS to be an improvement.

A larger, complete heatsink might be a solution for certain low power chips where the chip is fairly low power and the heatsink performs better than anyone could need in even the worse high-temperature environments, but for the typical CPUs you're talking about, it's better to have the versatility of choosing the heatsink size, shape, tech, and cost.

I'm more annoyed when Intel designs a heatsink so the fan clips on via a shroud it's built into so you can't just strap a standard full frame fan on instead. At least they usually used high quality fans instead of the thin and/or short lifespan sleeve bearing junk that even some supposed-major heatsink manufacturer used.

Another benefit to someone like me is that since the heatsink is so easily removed, I can use it for other purposes without much hassle. For example several times I have converted incandescent flashlights to LED by putting a chunk off an old heatsink in the head of the light.

Granted this is a pretty niche application as I only do this to flashlights with some redeeming quality to their design to make it worth the bother but never the less, modularity of a separate heatsink, or separate anything for that matter, tends to increase the flexibility in use of components.

Other times I have reused CPU 'sinks on budget grade video cards which had such a tiny 'sink with high RPM fan that it was obvious that would be barey adequate and the fan short lived, but then they take up an additional motherboard slot like the higher end cards do.

I have never had a problem keeping an Intel or AMD cool enough at stock speed. The industry did have to wake up to the increased demand of heatsinks around the Athlon XP and Pentium 4 era, but responded eventually in producing larger designs, and many with heatpipes in them... of course this was a chicken and egg scenario because formerly motherboards did not have a way to mount them, you need holes and sometimes even a bracket to take the weight or else you've produced a Very Fragile Thing.

Lastly, how would you intend to mount something like you've pictured, where you can't see to install the CPU in the socket, can't latch it down, because there's a permanently attached heatsink in the way? I'm sure *something* could be designed to accomplish it, but again I don't think the majority of CPU customers want to pay more for this.

 

[Confirmation]

You could randomly pick any product, in computing or not, and state there is some idealized way they could make it to where it performs better or lasts longer, but there is the cost to do so, usually some trade off, ultimately the question becomes Is it fit for its purpose?

Many times (far too many to count) I have looked at some product failure and thought to myself, if only they had spent 20 cents more on a better spring, or $1 more on a better connector, or whatever the earliest failure point was, could have been delayed with better design and yet if you multiply that level of design times the # of components in some complex products, they would soon become much more expensive, especially very complex things like automobiles.

I have always felt that some AMD and Intel heatsinks were marginal performers, but they do their job and an enthusiast knows you can buy aftermarket 'sinks. I have not only had no problem keeping CPUs cool with aftermarket 'sinks, I've even been able to overvolt and overclock them which is going beyond Fit For Purpose.

The vast majority of Intel's market has no problem with how the IHS is involved so there is no reason for them to change it. They do not want the price of CPUs to rise in order to meet your idealized construction methods.

Are you old enough that you recall that for a period, both AMD and Intel had open die CPUs without a IHS? These CPUs were more prone to damage from a ham-fisted heatsink installer, it was more important that the bottom of the heatsink be near perfectly flat (most were not at the time and cheap heatsinks today still aren't), and many consider the move back to an IHS to be an improvement.

A larger, complete heatsink might be a solution for certain low power chips where the chip is fairly low power and the heatsink performs better than anyone could need in even the worse high-temperature environments, but for the typical CPUs you're talking about, it's better to have the versatility of choosing the heatsink size, shape, tech, and cost.

I'm more annoyed when Intel designs a heatsink so the fan clips on via a shroud it's built into so you can't just strap a standard full frame fan on instead. At least they usually used high quality fans instead of the thin and/or short lifespan sleeve bearing junk that even some supposed-major heatsink manufacturer used.

Another benefit to someone like me is that since the heatsink is so easily removed, I can use it for other purposes without much hassle. For example several times I have converted incandescent flashlights to LED by putting a chunk off an old heatsink in the head of the light.

Granted this is a pretty niche application as I only do this to flashlights with some redeeming quality to their design to make it worth the bother but never the less, modularity of a separate heatsink, or separate anything for that matter, tends to increase the flexibility in use of components.

Other times I have reused CPU 'sinks on budget grade video cards which had such a tiny 'sink with high RPM fan that it was obvious that would be barey adequate and the fan short lived, but then they take up an additional motherboard slot like the higher end cards do.

I have never had a problem keeping an Intel or AMD cool enough at stock speed. The industry did have to wake up to the increased demand of heatsinks around the Athlon XP and Pentium 4 era, but responded eventually in producing larger designs, and many with heatpipes in them... of course this was a chicken and egg scenario because formerly motherboards did not have a way to mount them, you need holes and sometimes even a bracket to take the weight or else you've produced a Very Fragile Thing.

Lastly, how would you intend to mount something like you've pictured, where you can't see to install the CPU in the socket, can't latch it down, because there's a permanently attached heatsink in the way? I'm sure *something* could be designed to accomplish it, but again I don't think the majority of CPU customers want to pay more for this.

i found a solution for intel stock cooling and same amd, just ADD a new CPU FAN, 2 in the same way, and temps will go lower haha..

Are you old enough that you recall that for a period, both AMD and Intel had open die CPUs without a IHS? These CPUs were more prone to damage from a ham-fisted heatsink installer, it was more important that the bottom of the heatsink be near perfectly flat (most were not at the time and cheap heatsinks today still aren't), and many consider the move back to an IHS to be an improvement.

yeah, i remember 1990 computers, thats why IHS was done, but watching videos on youtube, many times intel sells the CPU's with a wrong thermalcompound inside, and users from other countries cant send it to the warranty, then they have to make a DELID to solve it, and many times the delid is imposible and they break the CPU or something small inside

i know if you quit the IHS, and for a mistake you press too much the CPU, you can damage it in a way, that you will start the computer and soon, it will get stucked, not bsod, just stucked, a cpu fail due damage, that happened with the PRESSCOT CPU

presscot in 2004 was a temperature cancer, then many people to solved it, tryied to delid it, or test many ways to fix the high temps, and a lot of cpu dieyed, but, presscot was good in winter season... nice temps in the room haha, loved presscot for that

Lastly, how would you intend to mount something like you've pictured, where you can't see to install the CPU in the socket, can't latch it down, because there's a permanently attached heatsink in the way? I'm sure *something* could be designed to accomplish it, but again I don't think the majority of CPU customers want to pay more for this.

i was thinking in something in the CPU and IHS, that when you adjust it, has a maximum and exact protection so that you cant adjust it more ( to protect the cpu and contact with the ihs )

as if you had to turn the screw 8 times with a reference that marks you that is adjusted at max, so, the user could easily change the thermal compound that comes between the ihs and the cpu, without the necesity to run the risk that comes badly from factory ( thermal compund), as happens to many people who dont understand why do they have 3 cores in 80 degrees and another in 50, that's because the thermal compound between the IHS and the CPU came badly placed.

 

[mindless1]

I'm not seeing these events of 3 cores 30' warmer than the 4th, except in cases where whatever was stressing it, wasn't stressing the 4th core as much.

If at stock speed and equal load, I had 80'C cores, let alone 90'C as you posed previously, I'd return the CPU as defective, get a replacement that didn't have this problem.

It wouldn't be that the cores are at dissimilar temperatures, it'd be that they're at 80'C or higher at stock speed. This is an unacceptable margin that's only going to get worse as ambient temp rises, dust starts to build up, etc.

Suppose it was a thermal interface defect. I feel Intel has some idea how to run their business better than anyone else does, that they feel it is more cost effective to replace the rare shipping defects once in a while.

Considering all possible problems with PCs, I don't see any point in focusing on this topic, it seems to have a much smaller impact than myriad other PC issues.

However you could always send your ideas to Intel or patent them in case there comes a day when they decide that's the way to do it.

 

[Confirmation]

I'm not seeing these events of 3 cores 30' warmer than the 4th, except in cases where whatever was stressing it, wasn't stressing the 4th core as much.

If at stock speed and equal load, I had 80'C cores, let alone 90'C as you posed previously, I'd return the CPU as defective, get a replacement that didn't have this problem.

It wouldn't be that the cores are at dissimilar temperatures, it'd be that they're at 80'C or higher at stock speed. This is an unacceptable margin that's only going to get worse as ambient temp rises, dust starts to build up, etc.

Suppose it was a thermal interface defect. I feel Intel has some idea how to run their business better than anyone else does, that they feel it is more cost effective to replace the rare shipping defects once in a while.

Considering all possible problems with PCs, I don't see any point in focusing on this topic, it seems to have a much smaller impact than myriad other PC issues.

However you could always send your ideas to Intel or patent them in case there comes a day when they decide that's the way to do it.

problem with warranty, is for example, in my country, due " the laws " it takes 3 months to send a packet, and 3 months to receive it, then, is impossible apply a warranty, then the normal users wont do nothing, and experienced users will try to dismount the ihs, but if only the ihs were easy to dismount, without have to buy something new, or intel include in his expensive product, at last, they could include an IHS remover or something, but they dont care and dont think in that things because they have too much trust in themselves i guess thinking their product are perfect or something i dont know, i just know that AMD is thikning more in the people, than what intel does, for that reason as a intel buyer, i started to change my mindway, for now, im sure my next build will be AMD, not intel, i was surprised buying an APU Build in 2014