Intel tells Core i7-7700K owners to stop overclocking

[Prysin]

On 2017-5-28 at 3:52 PM, EinZerstorer said:

Total bullshit. 

yes, your post is just that.

[potoooooooo]

On 5/7/2017 at 3:51 PM, Mr.Meerkat said:

A great example of this is the infamous YLOD and RROD  

 

Edit: As in the solder fracturing, not how heat transfer capability is reduced. 

That's not the cause, iirc they were caused by the system heating up so much that the soldered BGA cpu would get hot enough on the substrate that the solder holding it in place would melt and it would move a bit because nothing holds a BGA in place other than the solder

[JuNex03]

7 hours ago, potoooooooo said:

That's not the cause, iirc they were caused by the system heating up so much that the soldered BGA cpu would get hot enough on the substrate that the solder holding it in place would melt and it would move a bit because nothing holds a BGA in place other than the solder

I did not get the past generation of consoles so I'm a bit ignorant about this. But, am I understanding this correctly? The systems get so hot that it cooks the solder?

[SpaceGhostC2C]

8 hours ago, potoooooooo said:

That's not the cause, iirc they were caused by the system heating up so much that the soldered BGA cpu would get hot enough on the substrate that the solder holding it in place would melt and it would move a bit because nothing holds a BGA in place other than the solder

 

1 hour ago, JuNex03 said:

I did not get the past generation of consoles so I'm a bit ignorant about this. But, am I understanding this correctly? The systems get so hot that it cooks the solder?

It doesn't. The BGA problem was /is that the solder points get   fractured by repeated songs in temperatures. Precisely because it doesn't melt, the cracks don't fuse again, and are-soldering process is needed. 

 

You can watch the video Linus did with Louis (Lewis?) Rossman, where he explains how the BGAs get damaged, how the system (or an oven) doesn't generate enough heat to melt it, and how the machines that do actually re-solder them work. 

[Notional]

On 4/5/2017 at 11:31 PM, NumLock21 said:

http://overclocking.guide/the-truth-about-cpu-soldering/

 

What? So they can only get micro fractures by cooling the CPU to -55c? That's colder than the coldest place on earth, and way below what you can achieve without LN2. And then rapidly heating it to 125c? How is it even possible to run a CPU at 125c today? It will throttle to a stand still before that happens. This nonsense is busted.

 

I haven't seen any old Intel chip that was soldered, having any life span issues. Not that anyone would ever use a CPU until it's EOL. Utter nonsense.

[PCGuy_5960]

On 5/28/2017 at 4:52 PM, EinZerstorer said:

Total bullshit. 

Go apply for a job at Intel! They could definitely use your expertise! You clearly know more than their engineers do! 

3 hours ago, Notional said:

What? So they can only get micro fractures by cooling the CPU to -55c? That's colder than the coldest place on earth, and way below what you can achieve without LN2. And then rapidly heating it to 125c? How is it even possible to run a CPU at 125c today? It will throttle to a stand still before that happens. This nonsense is busted.

Did it occur to you that this thermal cycling happens at the factory? 

3 hours ago, Notional said:

Not that anyone would ever use a CPU until it's EOL. Utter nonsense.

Yet, you are using an EOL CPU

http://ark.intel.com/products/80807/Intel-Core-i7-4790K-Processor-8M-Cache-up-to-4_40-GHz

 

Guys, seriously, stop fucking thinking that you know more than Intel's engineers or der8auer (In case you missed it the article is written by der8auer) If you think that you know more, apply for a job at Intel. They could definitely learn a thing or two from you.

[Notional]

3 hours ago, PCGuy_5960 said:

Go apply for a job at Intel! They could definitely use your expertise! You clearly know more than their engineers do! 

Did it occur to you that this thermal cycling happens at the factory? 

Yet, you are using an EOL CPU

http://ark.intel.com/products/80807/Intel-Core-i7-4790K-Processor-8M-Cache-up-to-4_40-GHz

 

Guys, seriously, stop fucking thinking that you know more than Intel's engineers or der8auer (In case you missed it the article is written by der8auer) If you think that you know more, apply for a job at Intel. They could definitely learn a thing or two from you.

 

No. That thermal cycling doesn't happen at the factory. Why on earth would you cool down a CPU to -55c/-67f during manufacturing? Even so, the micro fractures only happens after ~300 thermal cycles of -55c to 125c. You would never see either of those extremes under any normal use.

Der8auer will experience things like this with LN2 cooling, MAYBE. That that is an irrelevant use case for all consumers.

 

EOL on the market sure, but not practical EOL. The product still works and will for years.

 

Besides, since we moved to solder without lead in it, you will see microfractures in the BGA before any of this would become an issue anyways.

 

Intel has chosen this move to save money and increase profit margins. Now, sure the dies might have to have a certain size for it to work properly. But the fact that Intel is selling you tiny dies for a lot of money, just shows how much of a ripoff Intel has become.

[potoooooooo]

11 hours ago, JuNex03 said:

I did not get the past generation of consoles so I'm a bit ignorant about this. But, am I understanding this correctly? The systems get so hot that it cooks the solder?

Correct. It would end up moving tiny bits at a time until eventually something lost contact and it stopped functioning.

 

There is no disadvantage to soldering IHS' because something like that won't happen, because it's only between the IHS and the die, it's not there for conductivity reasons