Does overclocking reduce life span/damage cpu

[Schmuckley]

 

[cadaveca]

Exactly. Even pushing them that far too, temperatures were in spec as well. I have seen guys go as far as 1.88V with DDR4 that is said to run at 1.35V maximum, and they have survived to run again. Although the same guys doing such things don't really care if something blows up in their face and needs replaced.

I hate to ask this, but why would you care? LOL.

I heard it mentioned by someone that first-run DDR4 was more like DDR3 than DDR4 fully, robust-ness-wise, so pushing 1.65V wasn't that big of a deal. I have also noted that DDR4 isn't like those hypers... those hypers kept the scaling going with voltage increases every step of the way, but DDR4 seems to drop off. At least, under air.

 

[sneekypeet]

I hate to ask this, but why would you care? LOL.

I heard it mentioned by someone that first-run DDR4 was more like DDR3 than DDR4 fully, robust-ness-wise, so pushing 1.65V wasn't that big of a deal. I have also noted that DDR4 isn't like those hypers... those hypers kept the scaling going with voltage increases every step of the way, but DDR4 seems to drop off. At least, under air.

I paid for the CPU (Don't really want to drop another $600 to replace it, and take a chance of getting a shit IMC), and isn't likely I would get another motherboard either. So there is a safety factor I have to consider. Also I am looking for something reasonable that I can say do this and that, and most users are warm and fuzzy with what I tell em to match my results.

If I were in your seat, I would care a little less

 

[Schmuckley]

Whatever that crappy ddr4 that I got is..it doesn't scale with voltage;pretty sure it's Microns.Will not budge over 2800.
I know it's straying off-topic,but does anybody know of any good kits?
PS:I heard Stella causes man-boobs.

 

[qubit]

NOt really, no, I don't think highly of myself at all. But I do enjoy a good joke, and because you missed it, you're miffed? Oh well. You can't please all the people all the time...


You're a pot. I'm a kettle. Who gives a fuk? only you it seems, I'm enjoying myself.

But really dude, google superconductors and inductance and you'll have your answer. I paid good money to go to school to learn this stuff. It makes more sense to NOT answer than it does TO answer. It's about being humble.


See, I don't care. You seem mad. Oh well. Why would my words get you mad? You soft or something? Here's a hint.. the sentence that preceded this one is a joke. Not a troll. Not me being glib... a joke. Laugh it off, move on. I can't help that my sense of humour annoys you.

And just to put it out there, i'm a 40-year-old jobless bum, living off my wife while still going to college, with a criminal record longer than the eye can see. I'm the last person who thinks highly of themselves. Doing reviews is meaningless. I do a poor job, and had W1zz hire someone to fix my poor writing. And this is something that makes me think highly of myself? WHUT!?!

You've missed the mark, mate.

Ok, we can't go on too long about this given sneeky's warning, but if you're being humorous can you please stick a smiley or two in it?! Otherwise it all comes out a bit serious and things can escalate with misunderstandings, lol.

Also, I think you may have taken my first post where I said about being pedantic as somehow negating what you said, but no, I wasn't. Being the proper techy nerd that I am, I love edge cases like this, so just wanted to talk about the one weird case where current exists without voltage. Oh and I'm hardly the last word on superconductors and I'm sure there's a lot to learn that I don't know. I don't think even scientists at the cutting edge of it fully get it afaik. If you have some interesting link about this subject, I'd love to see it (either here, pm, or a new thread).

I'm really sorry about you situation (bold bit) it sounds pretty tough buddy, but at least you've got a great wife who's supporting you. And no I'm truly not being sarcastic here. You wanna sound off in pm about it with me one day, I'm all ears.

I've never seen you as the bad guy and I'm more than happy to go back to being on good terms as of right now. This is water under the bridge for me now.

 

[Solaris17]

This o/c needs liquid cooling.

May i humbly suggest Stella Artois.

I like him ^ got a 12 pack in the fridge.

 

[sneekypeet]

Whatever that crappy ddr4 that I got is..it doesn't scale with voltage;pretty sure it's Microns.Will not budge over 2800.
I know it's straying off-topic,but does anybody know of any good kits?

It is off topic for this thread. PM me, I may be able to help you out.

 

[Ikaruga]

Obviously you've never met a stock-cooled 4770K. They throttle all the time, as do laptop chips. CPUs running 90c++ was common until we had thermal monitoring, and then people starting cooking eggs on their PCs, and temps all of a sudden became a thing that normal users though about. I have been OC'ing, really, since 586 days, when soldering was the way it was done.

I spent a good decade with overclocking, it was one of my hobbies, so yes I know what you are talking about, and I did not say that thermal throttle doesn't happen, all I said that you need better cooling if it does. I understand that throttle is part of the "normal" operation, but it's not part of the optimal usage and user experience, hence it's a good thing to avoid it, which can best achieved with better cooling (assuming voltage and clocks are already optimal).

 

[Aquinus]

so just wanted to talk about the one weird case where current exists without voltage

We could have an entire discussion on this topic alone. What you describe is a little unusual in a DC circuit though because even if voltage were to go to zero, the current induced by an inductor releasing its energy will result in a change in voltage since resistance is most likely going to remain constant. Since inductors resist change in current, the current created by the inductor will actually cause voltage as a result of the resistance in the circuit. Also zero resistance doesn't mean infinite current, it means zero resistance, as in zero energy lost as current travels through it. You're still limited by the number of electrons available to be moved (which is where Dave's usage of coulombs, which is the actual measurement of electric charge (not electric potential, volts,) comes into play.) Usually the case of current and no voltage is in AC circuits where inductive loads cause the current to go out of phase from the provided AC voltage and this is referred to as reactive power. This can cause a current draw when the AC circuit voltage is zero but, it can also be the case that a non-zero voltage can have zero current.

I just felt a little more information on the topic was in order as there seemed to be a bit of confusion. Not to say that what anyone said was particularly incorrect, just incomplete.

With that said, I would recommend playing with CircuitLab if it was still free. I'm pretty pissed off that it isn't anymore. It was a fun tool for prototyping simple circuits.

I apologize ahead of time if anyone considers this post off topic.

 

[dorsetknob]

With that said, I would recommend playing with CircuitLab if it was still free. I'm pretty pissed off that it isn't anymore

Funny that
It stopped being free after i Prototyped a T9000 CPU code name "Arni " on the Site
some one called Skynet sent me a DCMA notice

 

[Frick]

With that said, I would recommend playing with CircuitLab if it was still free. I'm pretty pissed off that it isn't anymore. It was a fun tool for prototyping simple circuits.

Oh good grief yes, that tool was a godsend for real. Now it's hateful, $16/month and you can't even use if commercially? It's like electrical drawings, your choices are an insanely complicated engineering CAD program or ANOTHER insanely complicated CAD program, and all you want to do is draw simple line diagrams and installations mortals without architecture/engineering degrees can understand.

 

[Red_Machine]

I have a 3770K OC'd to 4.5GHz, cooled by a Thermaltake Water 2.0 Extreme. Temps are so low, they're lower than the chip would be if it were running at stock clocks with stock cooling. I'm not worrying about degreadation at all, to be honest.

 

[cadaveca]

We could have an entire discussion on this topic alone.

Which, if you know a bit, exposes the means behind the method, given the stage this act has played out on, but it'd definitely make for a good subject in the science forum.


I have had many many CPUs degrade or die. The 4770K is used to review many Z87 and Z97 boards does not like to run 4 sticks over 2133 MHz any more, where before I was pushing 2800 MHz with ease (and all the way up to 3400 MHz).

The rest of the CPU's clocking ability, if I only use 2 sticks, works fine. I even just finished trying out my old Avexir 3100 MHz kit, and it still easily goes over 3300, and that's where I stopped playing.

I have both retail and ES 4790K chips. I have many ES chips, in fact. ES chips are very brittle in my experience, and cannot handle long-term OCs at all. If you veer outside of Intel's specifications, death approaches quickly, very similar to what other users report about degradation and needing more voltage. This isn't something that can be fixed by changing motherboards or any other component... I have enough hardware to be able to test for such things and have only had memory degrade, but almost never a motherboard unless it's been P67-based, or has had poor MOSFET cooling. Retail CPUs do not seem to suffer in the same way, but given that there are difference form one chip to the next, it's hard to ascertain what the real story is about how these chips degraded other than by OC'ing. And I can say that pretty specifically, because I quite purposely only OC within Intel's listed "safe maximums", which is far lower than you'll see most OC guides offer as OK.

 

[dorsetknob]

no one has mentioned the problem of electron migration in all this Kerfuffle

Given we are going to ever smaller die size and increased die density then electron migration is an ever increasing problem especialy if your overclocking these smaller and denser chips with higher Voltage and Current


Ps under an electron microscope damage from electron migration looks nearly identical to damage from ESD

 

[LightningJR]

no one has mentioned the problem of electron migration in all this Kerfuffle

Given we are going to ever smaller die size and increased die density then electron migration is an ever increasing problem especialy if your overclocking these smaller and denser chips with higher Voltage and Current


Ps under an electron microscope damage from electron migration looks nearly identical to damage from ESD

Indeed, this is also going to be a problem when trying to reach lower and lower process technology. We have already been seeing the effects of the limitations of going smaller when the trend of lower process isn't giving us much improvement in the way of power required or heat output.. At least in larger chips anyway.

We need a breakthrough in Graphene development that makes it more affordable to produce or some other material/idea. We'll see a massive slow down in Moore's Law if not, a Technological Depression...

 

[qubit]

We could have an entire discussion on this topic alone. What you describe is a little unusual in a DC circuit though because even if voltage were to go to zero, the current induced by an inductor releasing its energy will result in a change in voltage since resistance is most likely going to remain constant. Since inductors resist change in current, the current created by the inductor will actually cause voltage as a result of the resistance in the circuit. Also zero resistance doesn't mean infinite current, it means zero resistance, as in zero energy lost as current travels through it. You're still limited by the number of electrons available to be moved (which is where Dave's usage of coulombs, which is the actual measurement of electric charge (not electric potential, volts,) comes into play.) Usually the case of current and no voltage is in AC circuits where inductive loads cause the current to go out of phase from the provided AC voltage and this is referred to as reactive power. This can cause a current draw when the AC circuit voltage is zero but, it can also be the case that a non-zero voltage can have zero current.

I just felt a little more information on the topic was in order as there seemed to be a bit of confusion. Not to say that what anyone said was particularly incorrect, just incomplete.

With that said, I would recommend playing with CircuitLab if it was still free. I'm pretty pissed off that it isn't anymore. It was a fun tool for prototyping simple circuits.

I apologize ahead of time if anyone considers this post off topic.

Everybody, I've started a new thread about this, here: www.techpowerup.com/forums/threads/superconductors.214239