Holy Mutha of Intel CPUs!

[Virgorising]

Was just moved to search coolest running Intel desktop chips.....and LOOK WUT I FOUND!!!!!!! O M G.

http://www.makeuseof.com/tag/two-way...ing-intel-cpu/

I thought the NSA HDD firmware spy was the ultimate scandal.....but suddenly, here's a whole new one! WUT?

I think we humans need heatsinks more than our computers do now.

Could this story be true?

 

[videogames101]

4000 posts? hmm

 

[Virgorising]

4000 posts? hmm

This does not shock U?

I am totally shocked. TOTALLY. Perhaps it's a known phenomenon.....but it wasn't to me.

I get U R a gamer, but this is not virtual. And why you think you can determine anything via a given number of posts, I do not know.

 

[Cerb]

I don't recall any lying. We knew IB was not soldered from day 1, just like Haswell, and I can't recall ever hearing about any overheating CPUs, until the very recent issues with some i7-4790K samples, which may partly be motherboard/firmware related.

 

[Berliner]

Can someone sum this up?

 

[Virgorising]

Can someone sum this up?

U mean, determining its veracity without further delving? Following up?
_________________________________
Edit: maybe it's bogus. Or a promo for some upcoming The Real Housewives of Santa Clara show. But, if not bogus, wouldn't Intel sue? If I were Intel, I sure would.
________________________________
Another Edit: Been delving to try to chase accurate data. While it seems Intel was not lying about anything, the soldering really does shorten the life of the chip, tho. I have no clue if Haswells run very hot....never had one. BUT most humans do not upgrade their systems as immediately as it seems most members here do.

Having jus learned about this....I DON LIKE IT.

 

[Virgorising]

Am now confused delving. Does this mean Skywell S chips will all be BGA?

http://www.eteknix.com/intel-decided-to-stick-with-lga-platform-with-skylake-chips/

 

[Idontcare]

Was just moved to search coolest running Intel desktop chips.....and LOOK WUT I FOUND!!!!!!! O M G.

http://www.makeuseof.com/tag/two-way...ing-intel-cpu/

I thought the NSA HDD firmware spy was the ultimate scandal.....but suddenly, here's a whole new one! WUT?

I think we humans need heatsinks more than our computers do now.

Could this story be true?

That blog links back (references) to this forum and a thread I made.

The second theory claims that the distance between the die and the heat spreader can cause higher than normal temperatures in Ivy Bridge and Haswell. According to this theory, delidding removes 0.09 mm of glue between the IHS and the CPU’s circuit board; removing the glue shortens the distance that heat travels, resulting in lower temperatures.

^ the hyperlink takes you to one of my posts in the thread Delidded my i7-3770K, loaded temperatures drop by 20°C at 4.7GHz

 

[frozentundra123456]

Can someone sum this up?

Yea, a very strange post trying to make a big deal out of something that has been known for years.

 

[Ken g6]

One thing I hadn't seen before it's that the TIM might dry out in as little as two years. But TIM glued under a metal cap is probably going to last longer than TIM in the open under a heat sink.

 

[dave_the_nerd]

Could this story be true?

It is true that Intel uses thermal goop where they used to use solder, and it makes chips run hotter.

It is also not a big deal unless you're overclocking. Follow the published standards for cooling, etc., and you're good to go.

The idle and load temps on my 3570k haven't changed since I built the system 2 years ago. (They're higher than I'd like, but it's a cramped case so that's not unexpected.)

 

[flexy]

95% of PC owners, they don't even know what overclocking is. Temps are therefore not an issue for the majority of the market. Even for overclockers, temps are usually not an issue since you can get an aftermarket heatsink to keep temps resonable.

Was your post/link meant as sarcasm? Because that article is a total joke on so many levels.

Also..whether TIM dries out is entirely irrelevant. What counts is that the material which makes the TIM fills the micro-gaps between die and IHS/HSF for better thermal transfer. It doesn't play a major role whether a paste would actually really dry out as long as the material still fills in the gaps.

The rest is pure garbage tabloid "journalism", claiming something which I never heard of nor that there is any evidence for: That the TIM Intel is using is drying out (claim in the article) and therefore most Intel CPUs would die a heat death after some years. And then even implying it is done intentionally.

TOTAL GARBAGE ARTICLE.

 

[dave_the_nerd]

Was your post/link meant as sarcasm?

I think you need to grab a beer and then read OP's posting history. They're very passionate about this sort of thing.

 

[flexy]

>>
and it makes chips run hotter.
>>

I have a non delidded 4770k which runs CRAZILY HOT, but this only since I feed it with more voltage and I overclock. And the "crazy temps" show only up when I spend my time running Prime, XTU, OCCT etc..

Do you think that the "normal" user runs OCCT or Prime all day long?

It's funny because in real life apps/games etc., temps of i5/i7 are actually lower than previous gen CPUs...the entire TIM, delidding etc. issue is a non-issue for anyone BUT a handful of enthusiasts for whom temps under very unrealistic stress-testing plays a role.

 

[flexy]

I think you need to grab a beer and then read OP's posting history. They're very passionate about this sort of thing.

Phew (K I calm down now...sarcasm meter broke today or something )

 

[VirtualLarry]

The rest is pure garbage tabloid "journalism", claiming something which I never heard of nor that there is any evidence for: That the TIM Intel is using is drying out (claim in the article) and therefore most Intel CPUs would die a heat death after some years. And then even implying it is done intentionally.

TOTAL GARBAGE ARTICLE.

I tend to agree. I think that if Intel was producing chips that would experience "heat death" shortly after the warranty ran out, there would be an outrage, especially from businesses.

I know that the old "white paste" could dry out and harden (had to take a screwdriver and a hammer to a ceramic PPGA Pentium to remove the heatsink), but my experience with AS5, is that it doesn't need to be re-pasted for at least 5 years, if not longer. (I generally upgrade, or the fan on the heatsink dies, and I have to remove the heatsink before 5 years.)

It's my opinion that Intel's paste (Shin-Etsu?) is similarly durable.

Edit: Mods, can we get a title change on this thread, to something a little more accurate and relevant?
Something like "Intel IHS thermal paste quality on recent CPUs"? At least, something with a few less exclamation points.

 

[Burpo]

Yea, a very strange post trying to make a big deal out of something that has been known for years.

Yeah, really.. Welcome to 2012

 

[ninaholic37]

My Pentium M is still working great after 11 years. Glad I didn't upgrade

 

[Dufus]

Then I guess Intel messed up with mobile chips since they don't have an IHS.

Perhaps with the smaller die IHS soldering temperatures are unacceptable (too much heat transferred to the die) so a move to less efficient paste.

As for burning up, technically thermal control throttling should kick in to prevent that and one should be able to run a processor that is in spec at just below Tj target 24/7 for at least the guaranteed life of the CPU. <105C for 3770k providing it has a decent cooler to keep IHS below ~67C.

 

[Idontcare]

Could this story be true?

The short answer is no. The TIM Intel uses under the IHS to replace the solder they use to employ is not comprised of high-volatility compounds (the TIM is composed of very low vapor pressure components).

This is done to avoid "dry out" as a secondary measure, but primarily it is done to avoid thermal pumpout effects.

In short, if you have to make a choice in terms of whom you place your faith, I'd place your faith in the engineers at Intel over placing your faith in the random blog writings that can be googled.

You are getting what you pay for, and the blogs came for free. Think about that.

 

[flexy]

>>
producing chips that would experience "heat death" shortly after the warranty ran out
>>

You know, "out of the box", even the time where the CPU can turbo (the 'normal' turbo speed) is limited to a few seconds, along with other calculations that are done to keep the CPU constantly within TDP, thermal limits etc.

If you overclock you overwrite all those things of course in addition to supplying more voltage etc.

It's just not an issue, and (as said already) IRONICALLY, in 90% of "real life" circumstances those chips do actually run at "decent" temps. I know it's an oxymoron to say that, but it's just how it is. 75C in a game won't blow up the CPU or shorten its life. From that POV it's not even "less efficient" paste, it actually does its job like intended.

While thanks to the findings of idontcare etc. we KNOW that the high temps under stress are caused by that tiny gap between die and IHS...so it's silly to assume that a company like Intel (who SURE have experience making CPUs did some critical design flaw and then went through to produce the chips, regardless. (Especially a "flaw" they could have been corrected later on with Devil's Canyon but then still didn't because this gap of course still exists).

Also..I *think* we also already established it's not really the quality of TIM which is an issue here... the drying out of the TIM and subsequently failing of CPUs is nothing but a fantasy tale on that blog anyway.

 

[jihe]

Quick Doc, get back on your Delorean!

 

[DrMrLordX]

My Pentium M is still working great after 11 years. Glad I didn't upgrade

Well it's soldered of course. If it had that nasty TIM in there . . .

. . . it'd probably still be working.

 

[SPBHM]

"the revelation suggests that Intel CPUs may not last longer than three to five years before succumbing to overheating."

ivy bridge is almost 3 years old, be ready for all those failures starting soon

or not

 

[ShintaiDK]

This time warp thread shows exactly why people should stop believing in the techsite drama. Its gossip style that some sites/blogs specialize in.