EK Water Blocks Removes CPU Heat Spreader for Better Cooling of Ivy Bridge Chips

[N.Broekhuijsen]

EK Water Blocks has released a cheap kit that allows users of existing Supremacy CPU blocks to remove the IHS and let the water block make direct contact with the CPU die.

EK Water Blocks Removes CPU Heat Spreader for Better Cooling of Ivy Bridge Chips : Read more

 

[eklipz330]

"EK Water Blocks has released a cheap kit that allows users of existing Supremacy CPU blocks to remove the IHS and let the water block make direct contact with the CPU die."

NO. it DOESN'T 'ALLOW' you to remove the IHS. smh. i thought it was a kit that helped you remove it easily. i was interested in it until i realized this was just about freakin screws.

 

[warezme]

How is this scary or new? Back in the age of the Athlon FX and Pentium (heatburst) processors, that is how you connected your heatsink to the CPU, directly to the die. You had to make sure the paste was spread evenly and thinly across the die with a credit card and removed the plastic thingy off your huge copper heatsink with downward facing fan. It was an art to keep the heavy heatsink flat against the die without cracking it from to much pressure. As far as I'm concerned Intel can quit trying to make thier own solution and leave the die bare. I prefer bareback on my chips.

 

[offerings12]

LoL

 

[nforce4max]

I've done this my self but on much older builds and there is a risk of cracking the die especially for IB. For those who want to keep the ihs but need improvements use liquid metal alloy compound that is liquid at room temp.

 

[flyflinger]

April fools?

 

[willard]

[citation][nom]nforce4max[/nom]I've done this my self but on much older builds and there is a risk of cracking the die especially for IB. For those who want to keep the ihs but need improvements use liquid metal alloy compound that is liquid at room temp.[/citation]
Common misconception that you can get noticeably better thermal results from using a more expensive paste. There is about 5C difference between the best paste and worst paste, and anything that's at least as good as Arctic Silver 5 will be within one or two degrees of the top.

It's better, but it's not enough to actually buy you any more headroom.

 

[blazorthon]

[citation][nom]warezme[/nom]How is this scary or new? Back in the age of the Athlon FX and Pentium (heatburst) processors, that is how you connected your heatsink to the CPU, directly to the die. You had to make sure the paste was spread evenly and thinly across the die with a credit card and removed the plastic thingy off your huge copper heatsink with downward facing fan. It was an art to keep the heavy heatsink flat against the die without cracking it from to much pressure. As far as I'm concerned Intel can quit trying to make thier own solution and leave the die bare. I prefer bareback on my chips.[/citation]

Not always. All P4 systems with an LGA 775 P4 that I've seen use an IHS.

 

[blazorthon]

[citation][nom]willard[/nom]Common misconception that you can get noticeably better thermal results from using a more expensive paste. There is about 5C difference between the best paste and worst paste, and anything that's at least as good as Arctic Silver 5 will be within one or two degrees of the top.It's better, but it's not enough to actually buy you any more headroom.[/citation]

That is wrong. The paste used by Intel is utter garbage. Using even a decent past such as AC5 can improve heat transfer by around 20% according to tests done earlier (links to which are found in Tom's original Ivy Bridge launch article).

 

[InvalidError]

[citation][nom]warezme[/nom]How is this scary or new? Back in the age of the Athlon FX and Pentium (heatburst) processors, that is how you connected your heatsink to the CPU, directly to the die.[/citation]
Humm, no.

The dies on early Pentiums were mounted on the back of a ceramic slab to which pins were attached so the HSF was making contact with the CPU's packaging/substrate. This arrangement has rather high thermal resistance from die to heatsink.

When Intel went with FC-BGA, they introduced Slot-1 and AMD introduced Slot-A where the CPU ships with HSF bolted to the CPU+cache PCB.

The P3 is the only Intel chip I remember where end-users had to put their HSF directly on the exposed CPU die. Every other desktop Intel CPU after that (s370/P3) had the IHS.

 

[matt_b]

We shouldn't even have to resort to something like this. Why Intel has invested over the years all the money and R&D for their process of IHS and die attachment, and then go to TIM out of nowhere? Don't understand it and shame on Intel for doing this nonsense.

On another note, I've seen someone do this using a highly-conductive compound and lose about 20*C off of an OC already touching 80* C. That alone proved to me Intel's path moving to TIM and ditching their excellent flux-less solder method was the wrong direction.

 

[upgrade_1977]

Why put metal inbetween at all? They should make a block with an O ring (rectangle) type seal that fits around the top of the chip, and matched size hole under the cpu liquid cooler so the water can just flow onto the chip directly. Imagine the temps then.

 

[iam2thecrowe]

[citation][nom]InvalidError[/nom]Humm, no.The dies on early Pentiums were mounted on the back of a ceramic slab to which pins were attached so the HSF was making contact with the CPU's packaging/substrate. This arrangement has rather high thermal resistance from die to heatsink.When Intel went with FC-BGA, they introduced Slot-1 and AMD introduced Slot-A where the CPU ships with HSF bolted to the CPU+cache PCB.The P3 is the only Intel chip I remember where end-users had to put their HSF directly on the exposed CPU die. Every other desktop Intel CPU after that (s370/P3) had the IHS.[/citation]
all the amd duron and athlon xp range the heatsink was directly to die. quite a few intel mobile processors were (and probably still are) direct to die also.

 

[twelch82]

[citation][nom]iam2thecrowe[/nom]all the amd duron and athlon xp range the heatsink was directly to die. quite a few intel mobile processors were (and probably still are) direct to die also.[/citation]

Yep. That was the fun stuff with the Athlon and the clip-on heatsinks. You'd attach one side first, then had to push down with a lot of pressure to get the clip on the other side. The challenge was to keep the downward pressure balanced, even though the natural tendency is to push harder on the side that needs to clip in.

At least back then, you could buy a top-end Athlon for $100-150 though. Chip your Ivy Bridge, and you're looking at losing quite a bit more.

 

[InvalidError]

[citation][nom]iam2thecrowe[/nom]all the amd duron and athlon xp range the heatsink was directly to die. quite a few intel mobile processors were (and probably still are) direct to die also.[/citation]
I know AMD had many more exposed die models than Intel and that many of Intel's mobile parts also lack IHS. The comment I was replying to specifically named Pentiums and the only bare-core Pentiums with user-replaceable heatsinks are the P3s on s370.

In laptops and other highly proprietary form factors that may use the mobile variants, replacing the included custom HSF is usually impossible and in most cases, neither is overclocking. For those, whether there is an IHS or not makes no difference.

 

[digiex]

Once upon a time, AMD Athlon era, there was no heat spreader on the processor die.

 

[InvalidError]

[citation][nom]twelch82[/nom]At least back then, you could buy a top-end Athlon for $100-150 though. Chip your Ivy Bridge, and you're looking at losing quite a bit more.[/citation]
Back then, a 1GHz P3-Coppermine would have set you back ~$800 and that makes the $300 i7-3770 seem like a bargain today: 25-30X the processing power for less than half the price even before adjusting for inflation!

Personally, I am glad the days of ~$1000 "extreme" CPUs are over.

 

[boulbox]

They should really rename the title, it gives a false sense of entitlement. Man i was so looking forward to a simple and easy IHS mod.

 

[rantoc]

[citation][nom]warezme[/nom]How is this scary or new? Back in the age of the Athlon FX and Pentium (heatburst) processors, that is how you connected your heatsink to the CPU, directly to the die. You had to make sure the paste was spread evenly and thinly across the die with a credit card and removed the plastic thingy off your huge copper heatsink with downward facing fan. It was an art to keep the heavy heatsink flat against the die without cracking it from to much pressure. As far as I'm concerned Intel can quit trying to make thier own solution and leave the die bare. I prefer bareback on my chips.[/citation]

Or sanding the heat spreaders until you barely saw the die below, a fraction to much and the die was doomed!

 

[f-14]

[citation][nom]blazorthon[/nom]Not always. All P4 systems with an LGA 775 P4 that I've seen use an IHS.[/citation]

P4? seriously? how old are you?


i was thinking intel P3 FCPGA and Athlon XP processors. in fact it was my very first thought when i saw the chip " oh an amd Xp cpu".

i wasn't scared at all so long as the cpu cooler mounting system took most of the burden off the chip, which at that time was the socket itself.

this kit the first thing i would do is replace those washers with much wider ones to make up for the lack of structural support being placed on the MB. thickness and compression will be the 2 biggest concerns if you do this, so don't use rubber washers stick with hard plastic.

 

[blazorthon]

[quotemsg=10598484,0,361277][citation][nom]blazorthon[/nom]Not always. All P4 systems with an LGA 775 P4 that I've seen use an IHS.[/citation]

P4? seriously? how old are you?


i was thinking intel P3 FCPGA and Athlon XP processors. in fact it was my very first thought when i saw the chip " oh an amd Xp cpu".

i wasn't scared at all so long as the cpu cooler mounting system took most of the burden off the chip, which at that time was the socket itself.

this kit the first thing i would do is replace those washers with much wider ones to make up for the lack of structural support being placed on the MB. thickness and compression will be the 2 biggest concerns if you do this, so don't use rubber washers stick with hard plastic.[/quotemsg]

(heatburst) was stated after Pentium in the post I replied to. I assumed this was in reference to Netburst. Are you suggesting that there were Netburst P3s that I'm not aware of or that heatburst would be in reference ot something else?

Also, the post that I replied to said Athlon FX, not XP.

 

[InvalidError]

[quotemsg=10598484,0,361277]i wasn't scared at all so long as the cpu cooler mounting system took most of the burden off the chip, which at that time was the socket itself.[/quotemsg]
If by "burden" you mean contact pressure, you do want to put most of it on the CPU die itself on those pre-IHS CPUs and most earlier FCPGA HSF did not have any washers or anything.

The main reason washer or other stabilization mechanisms were added later during that phase was to reduce the likelihood that the HSF might tilt and grind/crush chip edges while the HSF is being mounted.

I'm just glad my P3 survived me putting a Golden Orb on it... scary sounds.