• Some users have recently had their accounts hijacked. It seems that the now defunct EVGA forums might have compromised your password there and seems many are using the same PW here. We would suggest you UPDATE YOUR PASSWORD and TURN ON 2FA for your account here to further secure it. None of the compromised accounts had 2FA turned on.
    Once you have enabled 2FA, your account will be updated soon to show a badge, letting other members know that you use 2FA to protect your account. This should be beneficial for everyone that uses FSFT.

burning in your cpu

C

ciM2pHat4U

n00b
Joined
Aug 16, 2004
Messages
60
Who here swears by burning in cpu (and memory, etc)? I have never done this before and I find it very interesting. I just built a new P4 2.8e computer and I've been using it for less than a week - is there a certain window of opportunity that the cpu (or ram) must be burned in in order to get the benefits (ie is it ever too late to burn in a cpu)? Where can I find some nice burning in tutorials?

Thanks in advance!
 
I think the term means that a CPU/RAM module doesn't perform at its maximum capability until it gets used for awhile.

This is the case with subwoofer speakers. After using them awhile they hit a little harder and such.

I don't know if there is scientific proof that this is the case with computer componants, nor for the speakers for that matter. I have heard it used pretty often when talkin about subs though.

OG
 
There are multiple defenitions of "burning in", one is what Opera said, and another is burning in your thermal paste. It does not reach its full conductivity until after 48 hours of use (or a number like that), so you should not overclock until about two days of running the cpu for best results.
 
The whole burning-in thing is rubbish, if you mean that letting a CPU sit at a certain clock will somehow increase its stability and overclockability. Just how does one prove this phenomenon? The answer is "you can't." If you have seen this through empirical observation, either your testing methods are inconsistent, or you have subtly changed operating parameters, whether it's a cooler ambient temperature or a shifting of the thermal compound slightly. There is no good justification for the CPU itself burning in, but I'll be glad to hear your evidence if you have it.

Long, exhaustive stress testing--that's good for determining the stability of an overclocked system. But make no mistake; stress testing does not "condition" your CPU.
 
Being solid state, a CPU (or memory, or a motherboard) that runs without fail for 48 hours should, in theory, run forever. A properly cooled and cared for CPU doesn't usually just "go bad" months after it's purchased, a CPU that intends to go bad will go bad right out of the box.

Burning in a part is good for showing a customer "look, I ran it for two straight days and it had no problems."
 
Operaghost said:
I think the term means that a CPU/RAM module doesn't perform at its maximum capability until it gets used for awhile.

This is the case with subwoofer speakers. After using them awhile they hit a little harder and such.

I don't know if there is scientific proof that this is the case with computer componants, nor for the speakers for that matter. I have heard it used pretty often when talkin about subs though.

OG
Click to expand...

It doesn't. When people burn in systems they stress test them to see if everything is in working order (based on the fact that most parts fail soon after they are installed. if at all). Some people think that they have to burn in their systems to get the best performance out of them cause of the wording used.
 
Its not something I believed in :cool:

I was simply offering an explanation of what a lot of people believe the term means.

I for one think games are some of the best stress tests for systems. Screw Prime95. If I can play my games for 6 hours straight every night without a hitch then my system is stable enough for me ;)
 
Well, if you have results of it, and you're selling a computer, then you cna tell the customer that it works at full load, and don't come bitching to me if it doesn't.

Well, not in that phrasing anyway.
 
Codegen said:
Well, if you have results of it, and you're selling a computer, then you cna tell the customer that it works at full load, and don't come bitching to me if it doesn't.

Well, not in that phrasing anyway.
Click to expand...
I dunno, I kinda like that phrasing... lol
 
Personally, I like to keep all new hardware under heavy load for a few days after I get it since it is most likely to fail when now. The sooner it fails (if it will) the easier it is to replace & the less likely that I've rolled the machine out to somebody who's started working with it.
 
I agree. I do a burn in, simply because if it is going to fail it will fail for me rather than my customer.

As for burn in for overclocking... screw that. Everything I've learned has pointed towards a finite lifespan, cranking the voltage and temp for a few days can only hurt that.
 
Yeah, so instead of your CPU life span being 10 years, it becomes 8. Being that most people upgrade every few years at the most, I don't see any reason to not overclock.

My motto is: Go [H]ard or go home!
 
Only reason I really ever burn-in is like mentioned, just to check system stability. I usually use Sisoft for that,and even when Sisoft passes Prime95 tends to fail, so yeah screw Prime95. I just run Sisoft for 24 hours and if it's fine after that, then I'm set. I saw someone say if they can run there games for 6 hours straight it's fine, but what if say one day you're in a diffrent situation? Maybe now you've installed winamp or something and now you're running some extra programs, system stability goes down now. So now while you're playing, having 5x damage in RA3, you get a random reboot without getting a screenshot first =(
 
Operaghost said:
Yeah, so instead of your CPU life span being 10 years, it becomes 8. Being that most people upgrade every few years at the most, I don't see any reason to not overclock.

My motto is: Go [H]ard or go home!
Click to expand...
No.

As many people would recommend in the past, crank the voltage (2v on a PIII or Athlon) and run it like that for a week, then start overclocking.

8 years my ass, you're lucky to get 8 months with that method. Kinda like reving a car over redline for 6 hours to "break it in".

That method of burn in is downright stupid.
 
I believe you misunderstood my comment. If you think increasing your voltage will give you an 8 month lifespan on your CPU, then I think you should ask around this forum about that. I'm sure theres umpteen dozen people who have disproved that theory.
 
Operaghost said:
I believe you misunderstood my comment. If you think increasing your voltage will give you an 8 month lifespan on your CPU
Click to expand...

He didn't really say that. At least not in that context.

"Burning-in" at one time ( but perhaps not in the same way now ) was a method described by Oldman above. In that right -out -of -the- box- people overvolted, sometimes to the extreme, then "added" to the stress that the cpu was already under in those conditions by then running "torture test" in Prime.

The belief was that this would somehow "condition" the cpu to run faster. "And the torture would make it talk" ;)

Ok, now whether or not that's even true, lets not argue and fight about it! That's simply one description of what "burning-in" means to some folks. Or did mean at one time. Or maybe it still does. Whatever? And the guy clearly stated that he was against "that" method of "burn-in".

But using "burn-in" tools like Sandra, Prime, etc. to test the stability or failure possibility of a new machine, is certainly a fine idea. Or to test a finely tuned and "thoughtful" overclock doesn't stink either. And Memtest for all occassions is simply great.

But I don't see where Oldman at anytime says overclocking is bad, or that he doesn't do it himself, or that he suggests that others don't do it, or that it will kill a cpu in 8 months.

So since he didn't "really say" any of those things, maybe he doesn't feel like "going home." And neither does the guy who doesn't overclock at all. Maybe he doesn't feel like it. Maybe he'll do it later.
 
You still don't understand what I'm saying. Increasing your CPU voltage isn't a big deal to a point, the common misconception of overvolting a CPU at default speeds to increase the overclock is just that, a misconception.

If you think increasing your voltage will give you an 8 month lifespan on your CPU, then I think you should ask around this forum about that. I'm sure theres umpteen dozen people who have disproved that theory.
Click to expand...

Ask around yourself, I'm sure there are quite a few people that will tell you overvolting a CPU is not a good idea, and the 8 month lifespan is dead on. The P4 northwood core is limited to 1.65v for a reason, going to 1.85 or 2v, for short term burn in even, is not going to extend your overclock. Voltage burn in, by the definition I speak of (overvolting the CPU by at least 33% at default speeds for an extended time period) is not a good idea. Cooking the silicon does not improve its overclocking capability and can often cause damage to lower an overclock, as well as the lifespan of a CPU.

I am not a prude when it comes to this, I've clocked the hell out of CPU's, I've bumped voltage on damn near everything to a point and I still do. Past that point, which is what I'm talking about, which was the voltage "burn in" at over recommended voltage for a specific core, is just stupid and has been disproven on numerous occasions.

Max voltage on the Northwood is 1.65 if memory serves, go ahead and ask about the lifespan of those that went over 1.7v. You'll get 50% that have no problems, you'll get 50% that have lost a CPU or two within a 12 month period.
 
Ya what he says!

My pardon for not seeing the gentleman answer for himself just moments after I "stuck my nose in." :eek:

P.S. I read this after a family party and therefor a bit tipsy, and I swore you said, ( I am not a "prune." ) And thought, why is that man calling himself a "prune" ? Eyesight is thankfully clear now.

But again, wholeheartedly agree with your statement.



0ldman said:
You still don't understand what I'm saying. Increasing your CPU voltage isn't a big deal to a point, the common misconception of overvolting a CPU at default speeds to increase the overclock is just that, a misconception.



Ask around yourself, I'm sure there are quite a few people that will tell you overvolting a CPU is not a good idea, and the 8 month lifespan is dead on. The P4 northwood core is limited to 1.65v for a reason, going to 1.85 or 2v, for short term burn in even, is not going to extend your overclock. Voltage burn in, by the definition I speak of (overvolting the CPU by at least 33% at default speeds for an extended time period) is not a good idea. Cooking the silicon does not improve its overclocking capability and can often cause damage to lower an overclock, as well as the lifespan of a CPU.

I am not a prude when it comes to this, I've clocked the hell out of CPU's, I've bumped voltage on damn near everything to a point and I still do. Past that point, which is what I'm talking about, which was the voltage "burn in" at over recommended voltage for a specific core, is just stupid and has been disproven on numerous occasions.

Max voltage on the Northwood is 1.65 if memory serves, go ahead and ask about the lifespan of those that went over 1.7v. You'll get 50% that have no problems, you'll get 50% that have lost a CPU or two within a 12 month period.
Click to expand...
 
I'd laugh at anyone who overvolted their cpu to try and burn it in. Driving your car down the highway for a thousand miles may be good for it, but that doesn't equate to more then running prime95 or something 100% cpu intensive for a day or so. You may try running your computer for a couple days straight before overclocking just to break it in, but there's really no evidence that any of this helps your hardware. You know for a fact it can hurt it, so I see little reason for anyone to even try this method.
 
Papa-Ming said:
P.S. I read this after a family party and therefor a bit tipsy, and I swore you said, ( I am not a "prune." ) And thought, why is that man calling himself a "prune" ? Eyesight is thankfully clear now..
Click to expand...
yea, I'm not a prune, but you were a bit pickled, hmm? :p
 
0ldman said:
yea, I'm not a prune, but you were a bit pickled, hmm? :p
Click to expand...

Yes, "pickled."

Face as red as a "beet".

Approching an age where I'll need "Prune Juice".

I could go on but then this may become a "salad thread".

Ah but then, the croutons would be nicely "burnt-in".

:D
 
There is definitely something called "burn in".

I was involved in system testing of telecommunications gear about 10 years ago. We ran the equipment "cold" for 24 hours, fixed any problems, then ran it in heat chambers for 24 hours at 60 degrees celcius. This was designed to "pre-age" the equipment to the equivalent of about 6 months. The engineering logic is that any new equipment will generally fail within the first six months, or it will last for 10 years+ . That means that solid state devices don't just either work or not work.

There is also the question of ESD (ElectroStatic Discharge). If handled incorrectly, ESD can cause the device (CPU, Memory, chips etc) to fail. However, what is not generally know it that the damage can partially damage the chip, and cause leakage between tracks in the chip. The chip can then work ok for a while, and then make the chip behave erratically for some weeks or months before it fails completely. Never handle chips without electrostatic plastic or wrist strap, or if not practical, don't touch the pins.


As for "burning in" CPUs, I doubt that burning in for a while increases the overclockability, however, it may halp the "bedding in" of the CPU/fan etc, as mentioned previously.
 
You must log in or register to reply here.
Back
Top