3770K De-lidded and lapped With Liquid Pro - Awesome results.

[dqniel]

Thanks guys. I didnt bother re securing the IHS. After 48 hrs it should be welded in place anyway. And its not like it moves about.

What I would fear, with not using an adhesive on the IHS and using something like Liquid Ultra or Liquid Pro, is having to one day remove the CPU from the motherboard. The Liquid Pro/Ultra will have solidified, but it doesn't form a very strong "weld." The weld will almost certainly break when the sliding motion of the retention clip occurs, and then you'll be left with an IHS and a CPU that has a rough layer of Liquid Pro/Ultra permanently affixed... with no safe way of getting it off to use the CPU again. I had to sand my IHS and HSF to get the stuff off so I could reinstall my CPU in a new motherboard.

 

[Stu @ MSD]

What I would fear, with not using an adhesive on the IHS and using something like Liquid Ultra or Liquid Pro, is having to one day remove the CPU from the motherboard. The Liquid Pro/Ultra will have solidified, but it doesn't form a very strong "weld." The weld will almost certainly break when the sliding motion of the retention clip occurs, and then you'll be left with an IHS and a CPU that has a rough layer of Liquid Pro/Ultra permanently affixed... with no safe way of getting it off to use the CPU again. I had to sand my IHS and HSF to get the stuff off so I could reinstall my CPU in a new motherboard.

Im not worried.
I could probably sellotape it back together and it would run cooller than it did from new. :biggrin:

 

[Idontcare]

What I would fear, with not using an adhesive on the IHS and using something like Liquid Ultra or Liquid Pro, is having to one day remove the CPU from the motherboard. The Liquid Pro/Ultra will have solidified, but it doesn't form a very strong "weld." The weld will almost certainly break when the sliding motion of the retention clip occurs, and then you'll be left with an IHS and a CPU that has a rough layer of Liquid Pro/Ultra permanently affixed... with no safe way of getting it off to use the CPU again. I had to sand my IHS and HSF to get the stuff off so I could reinstall my CPU in a new motherboard.

The truth of the matter is that you must approach delidding with a mind for practicality from the outset.

And part of the that mindset includes assuming you are not going to be using the chip for more than about 24 months in the first place.

Because if you truly intend to be using the same 3570k or 3770k 3yrs, or 4yrs, or 5yrs down the road from now then you absolutely positively have no business delidding it in the first place.

You should keep the stock engineered solution in place because you know it was designed to provide you with the very sort of operating lifespan you have in mind.

Delidding is an experience limited to the purview of those folks who don't mind burning their CPU's proverbial candle at both ends.

If the concern of complications that may come up down the road when relocating your CPU is enough to give you pause, then the truth is you should be giving pause to the thought of delidding in the first place.

 

[Stu @ MSD]

The truth of the matter is that you must approach delidding with a mind for practicality from the outset.

And part of the that mindset includes assuming you are not going to be using the chip for more than about 24 months in the first place.

Because if you truly intend to be using the same 3570k or 3770k 3yrs, or 4yrs, or 5yrs down the road from now then you absolutely positively have no business delidding it in the first place.

You should keep the stock engineered solution in place because you know it was designed to provide you with the very sort of operating lifespan you have in mind.

Delidding is an experience limited to the purview of those folks who don't mind burning their CPU's proverbial candle at both ends.

If the concern of complications that may come up down the road when relocating your CPU is enough to give you pause, then the truth is you should be giving pause to the thought of delidding in the first place.

Totally agree. Well put.

 

[dqniel]

If the concern of complications that may come up down the road when relocating your CPU is enough to give you pause, then the truth is you should be giving pause to the thought of delidding in the first place.

Obviously, I disagree entirely as I am somebody who delidded and also thought about future applications of my chip. Just because I wanted lower temperatures and was willing to void my warranty doesn't mean I wasn't thinking about future uses of my chip.

As with anything else overclocking related, practicality and planning don't have to go out the window

By delidding and then using adhesive I was able to get fantastic improvements in temperature under load and also retained my ability to swap my chip from one motherboard into another upon upgrading. Given that this is a real-world example of why using an adhesive is something to think about, I don't see why you're dismissing it entirely.

 

[Idontcare]

Obviously, I disagree entirely as I am somebody who delidded and also thought about future applications of my chip. Just because I wanted lower temperatures and was willing to void my warranty doesn't mean I wasn't thinking about future uses of my chip.

As with anything else overclocking related, practicality and planning don't have to go out the window

By delidding and then using adhesive I was able to get fantastic improvements in temperature under load and also retained my ability to swap my chip from one motherboard into another upon upgrading. Given that this is a real-world example of why using an adhesive is something to think about, I don't see why you're dismissing it entirely.

Why did you want lower temperatures? Intel already designed your chip to live a full and useful life with the stock CPU TIM.

 

[sm625]

It is probably only that concave when it is taken off. Surely it must be flat before delidding? It is amazing that simply replacing the TIM can make so much difference. I wonder if it is the TIM that is responsible, or is it just that they put too much adhesive around the outer rim, causing the IHS to be separated from the die?

 

[Idontcare]

It is probably only that concave when it is taken off. Surely it must be flat before delidding? It is amazing that simply replacing the TIM can make so much difference. I wonder if it is the TIM that is responsible, or is it just that they put too much adhesive around the outer rim, causing the IHS to be separated from the die?

It really is that concave.

The thing to keep in mind here is that it costs money to over-engineer things. It costs money upfront in the design and engineering phase of the project, and it costs more money throughout the production cycle to produce parts with stricter mechanical specifications.

The stock IHS, as non-flat as it is, the stock CPU TIM, as thick as it is, the stock HSF, as small and low-performing as it is...despite their drawbacks the one thing you can't say about them is you can't say they don't get the job done.

Despite all that IB doesn't have going for it you can still OC to 4.2GHz on all cores and be LinX stable.

Replace that stock HSF with high-end air and you can reach 4.7GHz without even modifying the IHS or the CPU TIM.

People who delid are doing so either to hit just a few notches higher in clockspeed or because they want lower temperatures while OC'ing for the sake of noise or power reduction (or they do it purely for hobby entertainment sake - because they can, so why not ).

 

[Stu @ MSD]

It is probably only that concave when it is taken off.

No mate, taking it off couldnt affect the surface shape.
The IHS is a pretty hefty and strong piece of copper. Its simply designed to a tollerance and that tollerance is the very reason we need tim in the first place.

In Intels defence, it does its job damn well. My chip ran 4.680ghz as standard, Folding 24/7. Its "Our requirements" that leave it found lacking, not Intels design.

 

[dqniel]

Why did you want lower temperatures? Intel already designed your chip to live a full and useful life with the stock CPU TIM.

To overclock higher.

 

[cytg111]

But, if i am not mistaken, even with the massive temp drop the OP got, he's chip is still not clocking higher. That makes those -whatever- temps they're operating at at stock pretty damn inconsequential.

 

[KingFatty]

People who delid are doing so either to hit just a few notches higher in clockspeed or because they want lower temperatures while OC'ing for the sake of noise or power reduction (or they do it purely for hobby entertainment sake - because they can, so why not ).

There seems to be the unstated assumption that de-lidding *necessarily* reduces the longevity of the chip. Can you elaborate on this?

What if a professional Intel engineer delidded - could he manage to do it in a way that did not reduce the life expectancy of the chip?

I keep thinking that if your chip is *properly* delidded with no mistakes, would the lower operating temperature extend its life? Does the reduced life expectancy assume the delidder will make at least some mistakes?

 

[Nemesis 1]

We have discovered a copper based product that doesn't solidifie. Been testing for 6 months . It works well so far but it has to be tested at least a year. If all works well we will jusr rebrand the product sell it as our own . As others do.

 

[dqniel]

But, if i am not mistaken, even with the massive temp drop the OP got, he's chip is still not clocking higher. That makes those -whatever- temps they're operating at at stock pretty damn inconsequential.

I can't speak for the results of others- only my own. I know that before delidding and using Liquid Ultra I couldn't find stability above 4.5Ghz, and I had to use my fans full-force (too loud for my taste) at that speed. I can now run 4.7Ghz with the fans' RPM running at a level that's comfortable.

Is there any practical reason/need for my computer to run at 4.7Ghz? No, but what practical use is there for most hobbies? Not much, outside of the happiness it provides for the hobbyist. Personally, I get great satisfaction out of knowing that my CPU is running faster, cooler, and quieter while consuming less power.

We have discovered a copper based product that doesn't solidifie. Been testing for 6 months . It works well so far but it has to be tested at least a year. If all works well we will jusr rebrand the product sell it as our own . As others do.

What is it?

I keep thinking that if your chip is *properly* delidded with no mistakes, would the lower operating temperature extend its life?

I assumed the same. I would actually be astounded if the chip I'm using only lasted two years. I've never used a chip that didn't outlast most of the components of my build, even with overvolting and overclocking... and I used to be a lot more reckless with regards to the overvolting. Anecdotal evidence, I know, but my personal experience obviously has an effect on my personal decisions with regards to what I'm comfortable modding on my computer.

 

[Stu @ MSD]

But, if i am not mistaken, even with the massive temp drop the OP got, he's chip is still not clocking higher. That makes those -whatever- temps they're operating at at stock pretty damn inconsequential.

Far from it mate, it's more stable and reliable under load, it's life expectancy is probably doubled, it uses less power and more interestingly, the IMC seems to have a new lease of life and is running my ram @ 2400mhz with pretty nice 10-10-10-25-1 timings. with my stable 105mhz bus and 4.7ghz this will wipe the floor with any 5.0ghz / 100mhz / 2133ram PC all day long and use less power doing so.

 

[cytg111]

Far from it mate, it's more stable and reliable under load, it's life expectancy is probably doubled, it uses less power and more interestingly, the IMC seems to have a new lease of life and is running my ram @ 2400mhz with pretty nice 10-10-10-25-1 timings. with my stable 105mhz bus and 4.7ghz this will wipe the floor with any 5.0ghz / 100mhz / 2133ram PC all day long and use less power doing so.

I understand what you're saying, and glad you're glad with your setup . Excellent, but "wipe the floor"? I dont see that happening at all here, of course, wiping the floor will mean different things to different ppl!
What i take away from these stories is that Intel did their homework when they decided on cooling, cause it really makes little difference in those temp ranges anyway, plus/minus 20-30 degrees.

 

[Stu @ MSD]

Excellent, but "wipe the floor"? I dont see that happening at all here, of course, wiping the floor will mean different things to different ppl!
What i take away from these stories is that Intel did their homework when they decided on cooling, cause it really makes little difference in those temp ranges anyway, plus/minus 20-30 degrees.

Ok, wipe the floor maybe extreme but in 95% of usage it will be faster for very obvious reasons.

As for the cooling, of course they did. It works admirably at what it was designed for and will last for years but for £20, why not improve on it? The less latent heat dumped into the system the better, and I bet you have spent far more than that on fancy lights for your fans? Lol

 

[cytg111]

Yea, I get that, but warrenty and other riscs included .. wouldnt be worth it to me . Estimating i'd need atleast a ~30% speedbump to warrent the procedure.

 

[dqniel]

I understand what you're saying, and glad you're glad with your setup . Excellent, but "wipe the floor"? I dont see that happening at all here, of course, wiping the floor will mean different things to different ppl!
What i take away from these stories is that Intel did their homework when they decided on cooling, cause it really makes little difference in those temp ranges anyway, plus/minus 20-30 degrees.

20-30C is little difference

 

[Stu @ MSD]

Yea, I get that, but warrenty and other riscs included .. wouldnt be worth it to me . Estimating i'd need atleast a ~30% speedbump to warrent the procedure.

We all have our own requirements and things that make us happy. Warranty doesn't bother me in the slightest. If it did I wouldn't have water cooled 3x brand new 7970's that cost me almost £1500. Lol

 

[Idontcare]

We all have our own requirements and things that make us happy. Warranty doesn't bother me in the slightest. If it did I wouldn't have water cooled 3x brand new 7970's that cost me almost £1500. Lol

When throwing caution to the wind, it is more a matter of personal preference than tangible risk reduction when we start picking and choosing our thresholds for folding versus anteing up.

Risky behavior is not converted into unrisky behavior simply because one takes a modicum measure of risk reduction at some point in the process.

In college I had friends who would smoke a pack a day but took up jogging because they were worried about their health. (true story!) If health really was the goal then quitting smoking would have been the action to take. But the delusion was there.

They wanted to engage in rewarding risky behavior (be it smoking or delidding) but they wanted to assuage their cognitive dissonance created by the fear of knowing they are intentionally putting something they value in a position of total risk of loss (be it their body/health or their CPU's longevity) so they do things that amount to token risk reduction measures like taking up jogging or gluing their delidded IHS back onto the CPU's PCB.

The benefits are nearly entirely psychological, they don't want to come to terms with the magnitude of risk they already brought upon themselves (or their CPU) with their activity, so they willingly underestimate the downside risks of their activities such that the upside potential of their token risk-reduction activity (jogging, gluing IHS) appears to deliver an outsized return on investment.

I don't hold it against any of them, it is the classic "can't see the forest for the trees" situation and you know you can do absolutely nothing to help them see the forest because they are so intensely focusing on the tree and aren't about to take a wider perspective of their decision making process (which is precisely what cognitive dissonance does to you, which is why it is so studied in the first place).

 

[dqniel]

When throwing caution to the wind, it is more a matter of personal preference than tangible risk reduction when we start picking and choosing our thresholds for folding versus anteing up.

Risky behavior is not converted into unrisky behavior simply because one takes a modicum measure of risk reduction at some point in the process.

In college I had friends who would smoke a pack a day but took up jogging because they were worried about their health. (true story!) If health really was the goal then quitting smoking would have been the action to take. But the delusion was there.

They wanted to engage in rewarding risky behavior (be it smoking or delidding) but they wanted to assuage their cognitive dissonance created by the fear of knowing they are intentionally putting something they value in a position of total risk of loss (be it their body/health or their CPU's longevity) so they do things that amount to token risk reduction measures like taking up jogging or gluing their delidded IHS back onto the CPU's PCB.

The benefits are nearly entirely psychological, they don't want to come to terms with the magnitude of risk they already brought upon themselves (or their CPU) with their activity, so they willingly underestimate the downside risks of their activities such that the upside potential of their token risk-reduction activity (jogging, gluing IHS) appears to deliver an outsized return on investment.

I don't hold it against any of them, it is the classic "can't see the forest for the trees" situation and you know you can do absolutely nothing to help them see the forest because they are so intensely focusing on the tree and aren't about to take a wider perspective of their decision making process (which is precisely what cognitive dissonance does to you, which is why it is so studied in the first place).

Comparing the adhesion of an IHS onto a CPU after delidding to taking up jogging after beginning to smoke is beyond a stretch. The act of smoking is in and of itself, unhealthy. Delidding a CPU is only "unhealthy" for the CPU if done improperly, until otherwise can be proven. As somebody else asked already- what, exactly, is going to shorten the lifespan of the CPU during the delidding process if completed successfully?

Adhering an IHS so that upon uninstallation/reinstallation there is less hassle adds very little cost and very little effort for gains that are tangible if the user foresees the need for swapping motherboards. A lot of enthusiasts (the type of people that will be delidding in the first place) tend to upgrade more often, so the prospect of replacing a motherboard within the lifespan of a delidded i5/i7 is realistic for many. If they don't plan on upgrading, they still have to factor in the risk of the motherboard failing before the CPU. That, obviously, would also end in the need for swapping the CPU from one motherboard to another.

If I had not adhered the IHS to my CPU after using Liquid Ultra between them, then when I changed from a Pro4-M to an Extreme6 a few months later there is a very high chance my IHS would have detached during the swap. The "weld" formed by the solidified Liquid Ultra between the IHS and the base of the HSF was extremely brittle and weak, and that's with a much larger contact area and both surfaces being copper. I would imagine that the "weld" between the IHS and the non-metal die is even weaker, and extremely easy to separate. Had the IHS and die separated during the swap, I would have had to lap the inside of the IHS and the die itself due to the solidification of the Liquid Ultra- no thanks. Adhering the IHS and to the CPU took $4 and ten minutes, cheaper and faster than lapping the inside of the IHS and the die and reapplying a TIM, and certainly less risky. Had I used IC Diamond instead of Liquid Ultra between the IHS and die, I probably wouldn't have worried about using an adhesive.

 

[VonDutch]

When throwing caution to the wind, it is more a matter of personal preference than tangible risk reduction when we start picking and choosing our thresholds for folding versus anteing up.

Risky behavior is not converted into unrisky behavior simply because one takes a modicum measure of risk reduction at some point in the process.

In college I had friends who would smoke a pack a day but took up jogging because they were worried about their health. (true story!) If health really was the goal then quitting smoking would have been the action to take. But the delusion was there.

They wanted to engage in rewarding risky behavior (be it smoking or delidding) but they wanted to assuage their cognitive dissonance created by the fear of knowing they are intentionally putting something they value in a position of total risk of loss (be it their body/health or their CPU's longevity) so they do things that amount to token risk reduction measures like taking up jogging or gluing their delidded IHS back onto the CPU's PCB.

The benefits are nearly entirely psychological, they don't want to come to terms with the magnitude of risk they already brought upon themselves (or their CPU) with their activity, so they willingly underestimate the downside risks of their activities such that the upside potential of their token risk-reduction activity (jogging, gluing IHS) appears to deliver an outsized return on investment.

I don't hold it against any of them, it is the classic "can't see the forest for the trees" situation and you know you can do absolutely nothing to help them see the forest because they are so intensely focusing on the tree and aren't about to take a wider perspective of their decision making process (which is precisely what cognitive dissonance does to you, which is why it is so studied in the first place).

then i see it like this,

if smoking is bad for your health, then its the smoking hot running ivy compared to that,
we delid, because we know its better for ivy's health,
like we stop with smoking cigarettes, we start trimming, thats delidding ,

i think the lower temps we have, thanks to delidding, when we oc ivy will give a better/longer live,
high temps are another reason cpu's degrade right?

i delidded mine because i couldnt even run prime at 4.5ghz,
i bought my 3770K to oc, was disappointent that i was so limited to do so,
not that 4.5ghz is a bad oc, its very nice actually to run 24/7,
but not if you hit 100+C..
after delid , i can run 4.8ghz as daily oc, without breaking a sweat with temps..
only limit i have now, is my Max Vcore, i can run IBT at 5.0ghz and hottest core is 84C,
thats with a simple aircooler


i was just thinking about leakage, as another reason to have low temps,
thats why i came here to ask actually ..lol
i wasnt really sure if leakage is bad or anything,
but if your ocing on air, and have very low temps, isnt that better, if you look at leakage?,
just wondering

 

[cytg111]

In college I had friends who would smoke a pack a day but took up jogging because they were worried about their health. (true story!) If health really was the goal then quitting smoking would have been the action to take. But the delusion was there..

- pretty darn insightful, IDC, didnt know we did that here ..

 

[cytg111]

Comparing the adhesion of an IHS onto a CPU after delidding to taking up jogging after beginning to smoke is beyond a stretch....

An analogy only works if you focus on the analog part, shift that and you move into erasmus montanus territory, and pretty soon your mom be flying (or sinking).