Buildzoid's 3700x Static Overclock Degrades Processor

[moproblems99]

A Ryzen 5 3600 allegedly should turbo up to 4.2 Ghz, which the one in those charts is not doing.

I'll leave this here:

 

[Zach_01]

A Ryzen 5 3600 allegedly should turbo up to 4.2 Ghz, which the one in those charts is not doing.
I guess I am doing fixed multi then.
It takes 1.46v at all-auto settings. Gets very hot under load, too.
I know this newer stuff can take more heat, but I'm from the days of keep it below 55c if it's AMD.

I’m sorry but you don’t know all of it. 4.2GHz on the 3600 is for single thread only and not all core. And it’s doing this opportunistically and not sustained. This means that it should not work or loaded to 4.2GHz on all core loads. One way or another if you do that you will exceed silicon limits. You know anything about this and the silicon FITness controller?

 

[Schmuckley]

I’m sorry but you don’t know all of it. 4.2GHz on the 3600 is for single thread only and not all core. And it’s doing this opportunistically and not sustained. This means that it should not work or loaded to 4.2GHz on all core loads. One way or another if you do that you will exceed silicon limits. You know anything about this and the silicon FITness controller?

Nope! Don't care, either. It's going to do what I want or I'll find something that will or a way to make it happen!

Tell me more about these silicon limits and electron migration.

I'm sure not finding out anything about this "silicon FITness controller " via search engine. Got a link?

I'll leave this here:

View attachment 146228

Or..amount of Vitamin V and temperature!
And how hard you kick it in the tail!

 

[Zach_01]

Nope! Don't care, either. It's going to do what I want or I'll find something that wil!

Tell me more about these silicon limits and electron migration.

I can’t care less of your mocking...
If you don’t know what your CPU is and what it can take that is your problem. Just don’t encourage others to do so.
You can keep your ideas and settings and I hope no one will be mislead by you.

But please do let us know when you are going to replace you CPU due to degradation.

 

[Bones]

I'll be honest - I HATE seeing clocks jumping around, drives me nuts seeing it do that so I set things for it to be at the speed I want period.

I can also speak from past experience that all this yo-yo'ing of clocks can affect a chip.
Mobile chips as an example do the same thing (Power saving features) but all this up and down of the core(s) does have a detrimental effect on the silicon over time, even if it's "Designed" for it which mobiles certainly are.

You can largely tell which ones were constantly boosting and which ones were not by the way they behaved when OC'ed, those that got boosted acted up and were less stable, ones that didn't have it going on all the time did better meaning the ones that got boosted more had more degregation to them.
Since silicon is the same material used in desktops AND mobiles it's only common sense these chips would be affected in the same basic way from the effects of all this boosting.

That's how I run mine and I ain't changing either, I do have chips well over a decade old now that still run about the same as they did when I first got them new as proof of it.

 

[Schmuckley]

I can’t care less of your mocking...
If you don’t know what your CPU is and what it can take that is your problem. Just don’t encourage others to do so.
You can keep your ideas and settings and I hope no one will be mislead by you.

But please do let us know when you are going to replace you CPU due to degradation.

Will do, Cap'n!
We'll see what happens.
Wager $20 it doesn't happen until next Ryzen release?

I'll be honest - I HATE seeing clocks jumping around, drives me nuts seeing it do that so I set things for it to be at the speed I want period.

I can also speak from past experience that all this yo-yo'ing of clocks can affect a chip.
Mobile chips as an example do the same thing (Power saving features) but all this up and down of the core(s) does have a detrimental effect on the silicon over time, even if it's "Designed" for it which mobiles certainly are.

You can largely tell which ones were constantly boosting and which ones were not by the way they behaved when OC'ed, those that got boosted acted up and were less stable, ones that didn't have it going on all the time did better meaning the ones that got boosted more had more degregation to them.
Since silicon is the same material used in desktops AND mobiles it's only common sense these chips would be affected in the same basic way from the effects of all this boosting.

That's how I run mine and I ain't changing either, I do have chips well over a decade old now that still run about the same as they did when I first got them new as proof of it.

I was getting instability with the clocks jumping around. Random reboots n stuff. Not now.
Voltage is still all over the place...no weird reboots, though.

 

[Zach_01]

I'll be honest - I HATE seeing clocks jumping around, drives me nuts seeing it do that so I set things for it to be at the speed I want period.

I can also speak from past experience that all this yo-yo'ing of clocks can affect a chip.
Mobile chips as an example do the same thing (Power saving features) but all this up and down of the core(s) does have a detrimental effect on the silicon over time, even if it's "Designed" for it which mobiles certainly are.

You can largely tell which ones were constantly boosting and which ones were not by the way they behaved when OC'ed, those that got boosted acted up and were less stable, ones that didn't have it going on all the time did better meaning the ones that got boosted more had more degregation to them.
Since silicon is the same material used in desktops AND mobiles it's only common sense these chips would be affected in the same basic way from the effects of all this boosting.

That's how I run mine and I ain't changing either, I do have chips well over a decade old now that still run about the same as they did when I first got them new as proof of it.

And why are we keep talking about irrelevant things? This is Ryzen 3000 topic thread about degradation and OC. What you know about past chips has 0 value or meaning. Still you can’t comprehend that static OC is not for 3000. Other CPUs are irrelevant in this topic.
This is not a general conversation about OC. Digest it...
Ryzen 3000 is for stock only usage. Unless you static OC it to 4.0GHz max.

I was getting instability with the clocks jumping around. Random reboots n stuff. Not now.

You are the only one experience this as far as I know. And it was full stock?
Nonsense...

 

[Schmuckley]

And we are we keep talking about irrelevant things? This is Ryzen 3000 topic thread about degradation and OC. What you know about past chips has 0 value or meaning. Still you can’t comprehend that static OC is not for 3000. Other CPUs are irrelevant in this topic.
This is not a general conversation about OC. Digest it...
Ryzen 3000 is for stock only usage. Unless you static OC it to 4.0GHz max.

This is not some new form of substrate, bubba. It's still silicon.
It's actually harder and more durable silicon that can take more heat.
And I say what people know about past chips IS relevant.

 

[Bones]

It is relevant since it's the same basic thing - Silicon.
Unless the composition of the chip is changed (Which in this case it's not) all the effects that goes with it remain whether it's a 3000, 2000 or whatever else. These effects of course will differ by chip as to how it gets used, that's just how it is and you attempting to dismiss it because "It not a 3000 Ryzen" changes nothing about it.

Silicon is still silicon after all these years.

 

[Zach_01]

This is not some new form of substrate, bubba. It's still silicon.
It's actually harder and more durable silicon that can take more heat.
And I say what people know about past chips IS relevant.
Also past MSI boards and AMD.. (snicker)

Keep it in plain English please. We are not all from US here. If we start calling whatever names this leads no where...

It is relevant since it's the same basic thing - Silicon.
Unless the composition of the chip is changed (Which in this case it's not) all the effects that goes with it remain whether it's a 3000, 2000 or whatever else. These effects of course will differ by chip as to how it gets used, that's just how it is and you attempting to dismiss it because "It not a 3000 Ryzen" changes nothing about it.

Silicon is still silicon after all these years.

Oh you both are going to know soon how different the 7nm is, the hard way...

 

[Schmuckley]

Keep it in plain English please. We are not all from US here. If we start calling whatever names this leads no where...


Oh you both are going to know soon how different the 7nm is, the hard way...

Hey, this chip's already made it through a board blowing up with the 12V GPU rail of the PSU going bye bye.
Yeah...don't use cheap fan controllers with Panaflos.
PSU still works if you wanna use a 5450.
CPU still works fine.

 

[Zach_01]

Oh that is your issue... having no fear is often lack of rationality.

Yeah...don't use cheap fan controllers with Panaflos.
PSU still works if you wanna use a 5450.
CPU still works fine.

And am I supposed to understand the meaning of this?

 

[Schmuckley]

Oh that is your issue... having no fear is often lack of rationality.



And am I supposed to understand the meaning of this?

Yes you are. It's plain English.
This 3600x has already been through a motherboard and PSU blowing up.

 

[Bones]

Keep it in plain English please. We are not all from US here. If we start calling whatever names this leads no where...

This I have to agree with, need to keep it civil.

Oh you both are going to know soon how different the 7nm is, the hard way...

I know 7nm is different, just because it's smaller still doesn't change the charateristics of silicon itself.
And do note I didn't just zap my CPU with voltage, I'm of the way of using less as in less is better here.

Let's get back to what this is supposted to be about - Degregation of a chip.

I know well about the effects along lines of going from larger to smaller nm , it can't take the strain as well as a chip of a larger nm make, there is no disputing that fact of it. Also bear in mind degregation speaks to "How well" the chip clocks up with a given amount of voltage over the course of time, Buildzoid's example shows the effects of this degregation at work and the effect(s) it has.

I can use something here as a way of thinking about it for example - Remember old light bulbs with a wire filament?

Thicker filaments can handle larger amounts of power, smaller ones cannot handle the same amount without burning out and this effect is really similar to what happens when you downsize the nm's of a chip.
That's also why voltage with each downsize in nm goes down, not up to get the same speeds from the silicon. Right now Ryzen 3000 is the smallest of these requiring the least amount of voltage to get to a given speed value, such as 4000MHz for example here.

Older chips being larger in size can take more voltage to get there without issue, newer ones cannot without burning out, just like a bulb's filamant would do based on it's thickness (Size). Efficiency and speed does go up with reduced size because there is less of it (material) to go through but at a cost of how it can handle the voltage if it stays the same, which it cannot and continue to work because it will burn out if voltage isn't reduced.

At least that's how I think of it.

And yes, being an OC'er after all this time I understand all too well about degregation, it's effects and how it applies to both, older and newer hardware.

Done with this - Carry on guys.

 

[Zach_01]

Yes you are. It's plain English.
This 3600x has already been through a motherboard and PSU blowing up.

That was not on my quote and question

Yeah...don't use cheap fan controllers with Panaflos.
PSU still works if you wanna use a 5450.
CPU still works fine.

This was...
What is the meaning of this?
Are you telling how you blow up your board?


Are you telling that because the CPU has survived this, then it can handle 1.4+V and static OC?
WoW... how wasn’t I thought of that!

 

[Waldorf]

@biffzinker
no, xfr and the like all off. just running PB (turns "on" when PBO is default/off).

@Bones/schmuckley
and how many cpus compared to global numbers?
how many have been running with "auto" settings in identical setup/environment to have a control group chip?
even if u had 100 chips in the past 20y would mean nothing compared to millions.
meaning you have ZERO proof, if we talk about ALL the cpus (of same brand/type/model).
so unless you guys setup multiple cpu the same exact way and put the same stress on them,
it means narda when it comes to "this works better".

same for different gens:
just because its an amd/zen, doesnt mean its the same or gets "treated" the same
and if all silicon is the same: how about taking a intel from 20y ago and run it like its a Zen cpu when it comes to voltages and clocks.
right.

 

[Schmuckley]

@biffzinker
no, xfr and the like all off. just running PB (turns "on" when PBO is default/off).

@Bones/schmuckley
and how many cpus compared to global numbers?
how many have been running with "auto" settings in identical setup/environment to have a control group chip?
even if u had 100 chips in the past 20y would mean nothing compared to millions.
meaning you have ZERO proof, if we talk about ALL the cpus (of same brand/type/model).
so unless you guys setup multiple cpu the same exact way and put the same stress on them,
it means narda when it comes to "this works better".

same for different gens:
just because its an amd/zen, doesnt mean its the same or gets "treated" the same
and if all silicon is the same: how about taking a intel from 20y ago and run it like its a Zen cpu when it comes to voltages and clocks.
right.

20 years ago, computer chips took 3v and ran from 66-160Mhz.
Also, AMD and Intel ran in the same socket.

 

[moproblems99]

I'll be honest - I HATE seeing clocks jumping around, drives me nuts seeing it do that so I set things for it to be at the speed I want period.

I can also speak from past experience that all this yo-yo'ing of clocks can affect a chip.
Mobile chips as an example do the same thing (Power saving features) but all this up and down of the core(s) does have a detrimental effect on the silicon over time, even if it's "Designed" for it which mobiles certainly are.

You can largely tell which ones were constantly boosting and which ones were not by the way they behaved when OC'ed, those that got boosted acted up and were less stable, ones that didn't have it going on all the time did better meaning the ones that got boosted more had more degregation to them.
Since silicon is the same material used in desktops AND mobiles it's only common sense these chips would be affected in the same basic way from the effects of all this boosting.

That's how I run mine and I ain't changing either, I do have chips well over a decade old now that still run about the same as they did when I first got them new as proof of it.

I mean to each their own, but I have laptops that will disagree with you. Sadly enough, none of us have a sample size big enough to claim anything. If you wanna run your chip at 24v, good on you. If someone else wants to run theirs at .5v, then good on them too. Just be mindful telling other people that this or that is bullet proof and gospel.

The fact remains, when you do stuff out of spec, the only thing that should be expected is problems. Whether or not you get them is a different story.

 

[eidairaman1]

Come on man... did you monitor his test bed?
If that was true be sure about it that he would know. He doesn’t relay on software monitoring.

Mtn is a troll same with schmuckley, just ignore them.

Bones is not Young.

Keep it in plain English please. We are not all from US here. If we start calling whatever names this leads no where...


Oh you both are going to know soon how different the 7nm is, the hard way...

@biffzinker
no, xfr and the like all off. just running PB (turns "on" when PBO is default/off).

@Bones/schmuckley
and how many cpus compared to global numbers?
how many have been running with "auto" settings in identical setup/environment to have a control group chip?
even if u had 100 chips in the past 20y would mean nothing compared to millions.
meaning you have ZERO proof, if we talk about ALL the cpus (of same brand/type/model).
so unless you guys setup multiple cpu the same exact way and put the same stress on them,
it means narda when it comes to "this works better".

same for different gens:
just because its an amd/zen, doesnt mean its the same or gets "treated" the same
and if all silicon is the same: how about taking a intel from 20y ago and run it like its a Zen cpu when it comes to voltages and clocks.
right.

Bones acknowleges the risks involved in overclocking so no need to worry about him.

Schmuckley on the other hand hasn't a clue with anything, and I agree his advice is not advice but rubbish. There is a reason I ignore certain members and communicate with others.

I mean to each their own, but I have laptops that will disagree with you. Sadly enough, none of us have a sample size big enough to claim anything. If you wanna run your chip at 24v, good on you. If someone else wants to run theirs at .5v, then good on them too. Just be mindful telling other people that this or that is bullet proof and gospel.

The fact remains, when you do stuff out of spec, the only thing that should be expected is problems. Whether or not you get them is a different story.

Bro overclocking is out of spec no matter how you slice it.

He is neither. He's literally made a media business out of overclocking, something many of us could never hope to do.

You are right, mtn is a AMD Troll and schmuckley are nonsense givers.

 

[Zach_01]

Bones acknowleges the risks involved in overclocking so no need to worry about him.

I figured that out, eventually...

 

[Vayra86]

@eidairaman1 Typical how the most clarifying post in the topic gets flagged Low Quality.

I respect @Zach_01 for his patience.. wow.

20 years ago, computer chips took 3v and ran from 66-160Mhz.
Also, AMD and Intel ran in the same socket.

They also didnt boost...

 

[freeagent]

Degradation is just fancy for breaking it in, just means you need to use more voltage after

20 years ago, computer chips took 3v and ran from 66-160Mhz.
Also, AMD and Intel ran in the same socket.

They still do I think, you just need a psu that’s on it’s way out, and a good hard stress test for the cpu. Hit go and you should see those rails nose dive. I’m referring to my x58 before I replaced the psu.
Also 20 years wasn’t that long ago

 

[Bones]

@biffzinker
no, xfr and the like all off. just running PB (turns "on" when PBO is default/off).

@Bones/schmuckley
and how many cpus compared to global numbers?
how many have been running with "auto" settings in identical setup/environment to have a control group chip?
even if u had 100 chips in the past 20y would mean nothing compared to millions.
meaning you have ZERO proof, if we talk about ALL the cpus (of same brand/type/model).
so unless you guys setup multiple cpu the same exact way and put the same stress on them,
it means narda when it comes to "this works better".

I hope you realize with this you just voided what the topic is about since even Buildzoid by this is but a singular example AND in this way, there is no way to provide such proof. There is no way ANY of us are going to have a lab setup with millions of chips (Test group(s) as you put it) lying around to test, run or anything else.

Even the post I quoted below says this and I bolded it for you.

We can only test what we have and go by the results we get - Just like Buildzoid did with his but I don't see many if any really arguing about what he came up with.


How many times does the same result from testing over and over have to play out before one can say it "Is" or at least probrably is?

At some point with any testing you have to come to a conclusion of some kind and say as much.
Certainly not in the millions as you make it out to be with your global test numbers - Even AMD themselves woudn't go that far and they make 'em, so I guess by your post we can discount what AMD themselves spec's for them too since they haven't had degregation testing done in batches by the literal millions in control groups.
They do test them to see if they work before shipping out and that's it.

BTW if you do have a way to test a global number of chips concerning degredation yourself, show us please along with your results from all this testing.

same for different gens:
just because its an amd/zen, doesnt mean its the same or gets "treated" the same
and if all silicon is the same: how about taking a intel from 20y ago and run it like its a Zen cpu when it comes to voltages and clocks.
right.

Reread my previous - I touched on that when I mentioned the lightbulb filamant comparison and did say things about differences when you reduce the nm size of a chip. Also reread another when I said I run mine with as little voltage as possible - I did say it because it's true.

I mean to each their own, but I have laptops that will disagree with you. Sadly enough, none of us have a sample size big enough to claim anything. If you wanna run your chip at 24v, good on you. If someone else wants to run theirs at .5v, then good on them too. Just be mindful telling other people that this or that is bullet proof and gospel.

The fact remains, when you do stuff out of spec, the only thing that should be expected is problems. Whether or not you get them is a different story.

Never made any claim what I said was "Bullet Proof" or gospel as you put it, results can and will vary.

Don't forget I'm at the bot too and "Out of spec" is the standing order of the day with chips frying and dying for me.

 

[freeagent]

Well to be fair, overclocking has always been "At Your Own Risk" since the very beginning.

If they tell you your warranty will be void if you do something they warn you not to do, that is essentially on you, for thinking you knew better.. regardless if they could tell or not. Chances are they can nowadays.

I'm not saying don't overclock, but be smart about it if you intend for it to last. I would imagine them to be fragile to a degree, look how tiny the process is. And jamming a bunch of vcore into it probably doesn't help much since it is nearly at its limit anyways.

 

[Bones]

Well to be fair, overclocking has always been "At Your Own Risk" since the very beginning.

If they tell you your warranty will be void if you do something they warn you not to do, that is essentially on you, for thinking you knew better.. regardless if they could tell or not. Chances are they can nowadays.

I'm not saying don't overclock, but be smart about it if you intend for it to last. I would imagine them to be fragile to a degree, look how tiny the process is. And jamming a bunch of vcore into it probably doesn't help much since it is nearly at its limit anyways.

I agree 100%.