i7 3rd generation faster than Ryzen9 3900x?

[Keullo-e]

The simply fact just is that high voltages at low current doesn't damage (= degrade) a CPU as AMD has designed they to work like that with their PBO auto-OC.

 

[mclaren85]

And which one of those fine people would you be? Since I asked about yours in relation to the issues you specify. Not "insert generic research with relevant keyword". I meant this issue, specifically, by you, your own. And it being your own I would imagine you would share that without going through a paywall to request access, since it would of course have been your own paper, or at least co-author.


ps. did you really read what you just linked? I got this inkling feeling you did in fact not now that I read through some of them because it is not what I asked about and I get the feeling you just went on a google search and hoped to find something fitting.

Ahh sorry I misunderstood. These are not my articles, I'm just an ordinary engineer who loves researching. I read much of the articles since I have access the ieee archive, but honestly not %100 of them. Only the relevant parts.
Please do your research too. You will surely find some interesting information.

Electromigration is a fact. And this fact is more probable in 7nm than 14nm.

 

[Deleted member 205776]

You know what else is a fact? That you can run a CPU at 1.5v on low current for 20 years and it won't give out. Getting real tired of this.

 

[maxfly]

Sometimes its best to simply giggle and walk away...

 

[freeagent]

Electromigration is a fact. And this fact is more probable in 7nm than 14nm.

If it ever becomes a problem, you could always use your warranty. I am pretty new to AMD, so I don't know exactly what I am talking about.. but the same guys who are telling you are the same guys who helped me. Experiment on your own.. set a fixed clock, and a fixed voltage, your FIT voltage so you don't hurt your CPU.. run p95 128k and let er rip. Should start seeing scary temps fairly quick. Go to bios set everything back to stock, heck even throw PBO on, save and exit. Run that same test and look at your temps, and amps. Should be lower, and nearly the same average speed as your fixed, if not higher over time.. I havent figured out how to get it to "auto" oc up to higher clocks than what PBO can do, but I can do fixed clocks

AMD did a pretty good job with these, they can oc better than me.. for now.

 

[claes]

@mclaren85

I’ve never owned a Ryzen CPU and have no experience with them, but I have to wonder — why not just try running it at stock? The processor has a warranty, and there are tens of thousands of them running out in the wild, and no mass recall, so, maybe, AMD’s engineers designed to processor the run in a specific way?

So far all of your research is theoretical — why not attempt something empirical? As @freeagent pointed out, you have nothing to lose — the processor has a warranty.

In the tradition of the scientific method, you could test at stock, which serves as a baseline to compare to thousands of other users, and then start capping the voltage, test again, turning off boost, test again, try both, test again...

Then you’ll have actually tested your hypothesis and, using system temperatures, can start to draw conclusions.

At this point, people running slower Ryzen processors with fewer cores are getting better scores than your superior CPU — there must be a mistake in your decision making somewhere...

 

[RJARRRPCGP]

Only Ryzen 5000 can comptete with Intel chips in this bench because of their IPC increase.

Makes me suspect it's latency-related, too, an Achilles heel of Matisse! Because that's where Vermeer excels.

 

[elghinnarisa]

Ahh sorry I misunderstood. These are not my articles, I'm just an ordinary engineer who loves researching. I read much of the articles since I have access the ieee archive, but honestly not %100 of them. Only the relevant parts.
Please do your research too. You will surely find some interesting information.

Electromigration is a fact. And this fact is more probable in 7nm than 14nm.

Funny, since most of the ones you linked are articles about new and (possibly) improved methodologies in measuring EM immortality, not the actual issue electromigration itself or how prevalent it is. How is this even remotely relevant to your CPU dying and how could you have possibly read them in that case?
With one of the longer articles being a whole 5 pages long yet somehow you didn't read it? I mean I could read them all in less than 15 min.

Otherwise, since you do have access to them and you did read them, please quote which part in 10.1109/VLSICircuits18222.2020.9162775 is relevant to the voltages relative longevity of your 3900x.

 

[freeagent]

Its gettin hot in here

 

[R0H1T]

Is this part of the regular troll thread of the week quota at TPU

 

[mclaren85]

Ok As I said earlier, I've made a mistake by claiming i7 is faster than ryzen because of wrongly chosen powerplan. (Ryzen- power saver while intel - high performance)
Well I also wanted to share the test results of intel under the power saver mode for those who are curious.
Result: Ryzen is faster than 3rd gen of intel even after core boost and pbo is disabled. Topic can be closed, thanks to all.