CPU BURN-IN

[pm]

Originally posted by: AckbarThere was a Intel guy at the beginning of the thread that said there's nothing to confirm or deny that this "burn-in" technique is true.

That's not exactly what I said. I said that:
"I am not saying that people who claim to see some speed-up from some form of burn-in type operation are lying or are wrong. Just that - based on my experience - what they are seeing is not coming from the CPU and that I'm not sure what they are seeing. "

Which is a bit different. I am positively confirming that CPU's do not improve over time. In fact, due to a variety of mechanisms, they gradually get slower over time. It doesn't matter what you set the voltage to, or how well you cool it (or heat it), over time it will gradually slow down. If there's something going on - against which I withhold judgement - then I have no idea what it could be, but it's not inside the CPU.

 

[chinkgai]

wana hear a funny story? so after i read this thread, i decided what the heck...im gona try this just for kicks!

so i decided to find the lowest voltage i can barely get into windows with...AND I HAD TROUBLE FINDING THAT VOLTAGE!

i previously thought that i was stable at 1.482v @ 2.4ghz with my 3400+

well after last nite...i had to go all the way down to 1.274v :Q just to NOT be able to get into windows! so i did the "conditioning" or whatever exactly 1.3v for like 12 hours

now i guess i gotta search for what voltage my cpu is REALLY stable at 2.4ghz (currently priming at 1.352v for over 1hr so far...)

EDIT: im not sayin that this process works...yet...since i was able to prime for 20 min before stopping it when i was looking for "rock bottom" BEFORE the conditioning

 

[Duvie]

It may somehow be conditioning the lines of power from the PSU or the mobo...Maybe it helps cut down on noise or spikes caused by the mobo...

I usually really work my cpus and components...I tried the method and I got zealch...I maybe lasted prime like 20minutes versus 10 minutes and then I tried some more and it went back to 10 minutes....

I agree with PM....he should know....

 

[VirtualLarry]

Originally posted by: Jeff7181
Until I see an explanation of WHY any type of "burn in" would do anything to increase overclockability, or allow lower voltages at the same speed I can't help dismissing all these "results" as coincidence.

I actually feel the same way about the "freezer trick" for HDs.
:thumbsup: for an objective look at "computer voodoo" incantations such as being discussed in this thread. I don't believe in speaker burn-in, per se, either, but due to that they have mechanical seams, they could be modified mechanically, slightly during the very first hours of use. But I don't agree that specific types of music would help it more, nor that any significantly-altered sound difference that becomes apparent after that period is a good thing. More than likely, the speaker got damaged instead.

 

[chinkgai]

k, tested and zero gains.

 

[WildHorse]

Originally posted by: brazzmunk
i've read couple threads about this and wanted to get more detail about this...
how does it work and what is the end result.. it seems all too good to be true (the CPU burn in that is)

In addition to everybody's comments above,
Anandtech has this FAQ article about cpu burn in.

 

[Zebo]

Seems like a lot of work for screen-shot olympics. If it does "work" it's a fractional, un-noticable, almost infinitesmal change for a sh1t pot of time relative to regular OCing which I can do in less than an hour for the most part.

I'm pretty much a safe OCer though compared to most, don't like additional volts, perfectly happy with the highest rated chip in an iteration from the lowset cost chip in that iteration..

There's is so many vairables that can account for this "effect" it's not funny.
1. Simple placebo effect
2. Temperature changes in the room enabling CPU to run slightly higher or lower MHz at same Vcore so close to edge of prime failure.
3. It takes 200 hours for thermal compouds to fully "cure". When exacty did you start you're "burn in" 2 hours after building you say, I see.
4. Mobo capacitors delivering more constant power as their chemistry "burns in" over time.

 

[lion1]

It may be more helpful to look at this from a different point of view:
overclocking is the quest to find where the hardware fails - and then run just under that.

My guess is that Intel/AMD rate their products at where they can run safely for a variety of different environments, with the largest possiblity that it will always run (that is, until it completely wears down). But not necessarily at the maximum speed it is capable of.

So, currently my A64 3200+ is perking along at 2480 MHz. It will die sooner than if I would have left it at the stock 2000MHz - but - in the meantime it does some tasks faster. Don't know if I can push this cpu further...

 

[exar333]

a good point that PM brings up is that he has never seen cpu's get ebtter over time...i am assuming they have very expensive testing platforms for cpus..one's that have been running and been used extensively....as stated numerous times, many people "burn-in" their cpu's in the first few weeks of operation...this would lead me to believe that the "burn-in" myth does seem to apply to the cpu, but to other components (MB, RAM, ect...)

BTW, great Thread! interesting and no one has degenerated to AMD burns better than Intel...blah blah blah...hehe

it's nice to have such informed ppl here such as PM who can give some real technical expertise to these discussions, his time is appreciated!

 

[Elcs]

Originally posted by: pm
I don't deny that something may be happening at the system level, my comments are regarding the CPU only. I have been working at a CPU designer for Intel for over 10 years starting with the Pentium and more recently on the Itanium family. I have spent most of my days in the last year determining the critical paths on a 90nm CPU ("critical paths" are the circuits that limit the frequency of a design) and finding fixes for them. I spend a good chunk of my day running shmoos (a 2D plot of voltage versus frequency used to visually help with determining circuit marginalities) and trying to isolate issues.

I have never seen a mechanism that causes a given CPU operating at a given V/T point to speed up occur on a statistically large sample of parts. Integrated circuits are fundamentally crystals that are as perfect as the manufacturer can make them when they come out of the fab. They are - to the limits of manufacturability - as perfect crystals as we can make with as few dislocations, voids, grain boundaries, vacancies and impurities as possible. Using these integrated circuits involves subjecting the crystals to sharp thermal changes as well as voltage changes that result in the crystal gradually degrading over time. Normal operation changes the crystalline structure over time for the worse. Carriers get stuck where you don't want them. Charge traps are formed at the boundary. High current wires that are thinned and under thermal stress thin even further and grain boundaries become more pronounced. Under normal operation, eventually the integrated circuit will stop working. IC's are not like cars or other things with mechanical parts. There is nothing to "wear in". They only wear out.

I am not saying that people who claim to see some speed-up from some form of burn-in type operation are lying or are wrong. Just that - based on my experience - what they are seeing is not coming from the CPU and that I'm not sure what they are seeing.

I had a lot of reservations about the overvolting advice that was going around the net several years back. The idea was to run the CPU at an elevated voltage for a few days, and then afterwards it would run faster - and this idea is just fundamentally wrong. There may be some truth to the idea of setting thermal compound - or maybe it sets the grease in the fan ball-bearings or something - but the idea that the CPU is running faster was just plain wrong. And the idea that you could bias a PMOS gate such that it is repelling electrons and then somehow have a bunch get trapped in the gate creating trapped negative charge carrier is so wrong that I could show anyone in this thread why it's crazy with a few diagrams.

I don't have any such reservations with the idea of undervolting. If people think that it helps - it might in some way that I'm not familiar with - then it's certainly not doing any harm. I don't pretend to know all of the answers - I can only relate my experience. I have never said that "burn-in" doesn't do anything - only that whatever is going on, it's external to the CPU.

It's also worth mentioning that it's easy to be fooled by 'noise'. A very small improvement, may not actually be an improvement at all. There's a great (great!) book called "Voodoo Science: the road from foolishness to fraud" by Robert L. Park which talks about people throughout history who have been fooled by the "placebo effect", and by people who have been fooled by "noise" into thinking that something exists where it isn't (cold fusion, for example). 25MHz, or even 50MHz is a very small amount of 2000MHz or 3000MHz. I have a hard time reliably and repeatably isolating my failures down at that resolution using testers, thermal systems and power supplies that cost in the high six figure range. Just as I try to keep an open-mind about improvements, it's worth people keeping an open mind regarding seeing possible low-value illusionary "signals" amongst the noise.

(link to "Voodoo Science" at Amazon.Com Check out the reviews at the bottom... I'm not the only person who really liked it. If you like reading well-written and entertaining books on science, this one is a winner)

Very detailed explanation. Whilst it might not do much to physically prove or disprove the theory, it does throw up some interesting insights into the insides of a processor and some nice basis for countering the theory of burn-in.

Im looking at this as a rather inexperienced overclocker. Only 2 prime stable overclocks under my belt. Im also not a scientist, nor do I have any grounding in the field beyond very basic stuff taught at school.

I feel like setting the thermal compound may be the major issue affecting this theory. Once it sets and gets to an optimum level, it does a better job at cooling and that may be the deciding factor in the overclock.

pm: You mention that the crystals in an integrated circuit are 'as perfect as a manufacturer can make them'. Is it possible that a clean burst of a higher voltage could clear up some sort of imperfections you talk about? My understanding of such things is the same as the average joe public but I can imagine voltage being water pushed down a hosepipe (the cpu) to try and clear a blockage.

This entire thread is interesting.

 

[pm]

pm: You mention that the crystals in an integrated circuit are 'as perfect as a manufacturer can make them'. Is it possible that a clean burst of a higher voltage could clear up some sort of imperfections you talk about? My understanding of such things is the same as the average joe public but I can imagine voltage being water pushed down a hosepipe (the cpu) to try and clear a blockage.

Manufacturers use a "high temperature anneal" step in the manufacturing flow to essentially do this. See this page and do a search on the word "anneal".

There is also a "burn-in" step in the manufacturing flow after the die are wafer tested, cut, and packaged. The chips are put into a high-temperature, high voltage environment and left there running for some time. (there's a good, short, wikipedia article on this subject here). Burn-in tries to eliminate reliability failures in the field (ie. in customers computers) by stress-testing the chip. This step actually results in a significant slow-down in the device frequency caused primarily by a reliability issue called BTI (bias-temperature instability) and to a lesser extent by NMOS hot-electron gate impact ionization. A lot of research has gone into trying to reduce this frequency degradation and Intel has published some interesting papers on the subject in recent years at reliability conferences.