Burn in week (my method)

[BarboBot]

Well after reading a lot of info on cpu burn in i've decided to burn mine in due to the increased performance on lower power. i went through and found the stock voltage for my 3200+ venice was running at 1.4v so i ran cpu burn in for 12 hours at the stock voltage (no problems) next i've moved it down to 1.36v i'm running another 12 hour burn in session then i'll move it down to 1.2v then another 12 hour session. I'm hoping to get increased stability with this gradual burn in process as opposed to trying to go straight to the lowest voltage possible. I'm also hoping i can get a little bit lower voltage than the traditional burn in process since i'll be gradually building up stability at lower voltages. I was just wondering if anyone else had any input or suggestions/opinion on this process?

 

[garkon]

Dose this work? Seems like it would not really affect stability too much, as i've never had stability problems w/ just installing OS updating all the way, running a anti vir, and keeping it pretty clean. Dose it have to do w/ overclocking? I just figured that if u were overclocking ud want to up voltage. I know nothing about overclocking, actually interested in an answer from someone who knows what there talking about!!

 

[Maximilian]

Originally posted by: garkon
Dose this work? Seems like it would not really affect stability too much, as i've never had stability problems w/ just installing OS updating all the way, running a anti vir, and keeping it pretty clean. Dose it have to do w/ overclocking? I just figured that if u were overclocking ud want to up voltage. I know nothing about overclocking, actually interested in an answer from someone who knows what there talking about!!

As am i. Ive never heard of this "burn in" thing before.

 

[imported_oralpain]

The only thing I have seen burining in help with is Winbond and older samsung memory.

I have never had burning in help a CPU.

Can never have enough honest experimentation however. Find the lowest voltage it can run stable at stock, then finsih your burn in and check again.

 

[Shimmishim]

yeah... there's no proof that it works though there are some good threads about it...

i've seen the results of burn in give me more stability in prime.. (meaning it ran longer) but no increase in overclock

 

[BarboBot]

its so you can overclock higher at lower or stock voltage. you're training you parts to run with stability at lower voltages, the lower the voltage you can run at stabily the less power consumption, also less heat is generated, i'm running an air cooled system and i'm into the second phase of this experiment and its already running cooler. (at stock volts 1.4 the cpu was running at 37-39c now its halfway through the 1.36 volt burn in session and its only 35-36c) so its running cooler also after this week is up i'm planning on running the test again at stock voltage the end result i'm looking for is higher over clock at stock voltage and lower temperatures under load. (ohh and on side note prior to this burn in my comp ran at 32c i checked it today and its down to 29c at idle) I've read many a articles about cpu burn in and there is sound proof of it working should you look into it. www.ocforums.com has a lot of info/guides on the topic. granted my method of burning in is slightly differs from the conventional method used i'm just trying it out to see what results i get (so far the results in my two days of this experimental burn in has yeilded positive results)

 

[imported_michaelpatrick33]

Originally posted by: BarboBot
its so you can overclock higher at lower or stock voltage. you're training you parts to run with stability at lower voltages, the lower the voltage you can run at stabily the less power consumption, also less heat is generated, i'm running an air cooled system and i'm into the second phase of this experiment and its already running cooler. (at stock volts 1.4 the cpu was running at 37-39c now its halfway through the 1.36 volt burn in session and its only 35-36c) so its running cooler also after this week is up i'm planning on running the test again at stock voltage the end result i'm looking for is higher over clock at stock voltage and lower temperatures under load. (ohh and on side note prior to this burn in my comp ran at 32c i checked it today and its down to 29c at idle) I've read many a articles about cpu burn in and there is sound proof of it working should you look into it. www.ocforums.com has a lot of info/guides on the topic. granted my method of burning in is slightly differs from the conventional method used i'm just trying it out to see what results i get (so far the results in my two days of this experimental burn in has yeilded positive results)

Are you using Arctic Silver? That does settle in and can drop your temps after a few days/week but either way good going.

 

[rise]

i'm trying this now barbo using sentinels method from xs. i just started and i'm actually in my first 8 hour run using burn-in 1.01.

i dunno how well it works but i'll report back to you. his method is slightly different as he starts at the highest stable oc and backs the volts down until it will boot and run prime for no more than 5 minutes. then you start the burn in there. as you can turn off the "error detection" in burn in it will just run, not like prime, for however long you specify.

once it burns in, run prime for a while to see any progress and repeat. lower volts til just barely stable and burn again.

from what i've read over there some people have had very good to no results. i'd like to get at least my 2750 to run at 1.45v and ideally i'd like 2850+ at 1.5v.

who knows? no harm in trying

 

[Vegitto]

It worked charms for my Athlon 1400+. My mobo doesn't have any locks, so this was a perfect solution .

 

[RichUK]

yeah there was loads of talk of this burn in before.. but alot of people have said that this doesn't really help the OC

 

[Vegitto]

Nope, I'm not sure it does. But it helps with the voltage. My mobo won't let me fiddle around with the voltages, so I just overclocked until unstable (which was really, really soon) and then burned in. It worked.

 

[BarboBot]

like Vegitto said its not to help my OC its to make it stable at lower voltages and it will it run a little cooler at the stock voltage (hopefully), i'm into my second burn (this one is only running 7hrs 30mins) comp booted up just fine at 1.34v so i went ahead and rebooted and started in on the next session. i'll drop it again and run another 8hr session tomorrow morning while i'm at work. its starting to run a little bit on the warm side so this'll be were there real gains will be made i think i'm hoping to get a successful 1.2v possibly 1.1 burn in done then i'll leave the cpu alone and move on to the ram and the gpu, but i'll keep posting the results of this experiment i've got going. and keep me posted Rise and good luck to you just be patient and take your time like you said no harm in tryin

 

[RichUK]

so burn in just allows you to run lower voltages, and not OC higher ... so the max OC could be like 2.6 @ 1.55, but after burnin you could get that down to 1.45 or summin but not run it any faster for example .. cuz if thats the case im going to do that on my next proc 4400+


EDIT: ahh just got your post in before mine BarboBot, anyways you answered my question

 

[Vegitto]

Try for 1V, after that, overclock a bit, and increase voltage minimally. Standardly, the 3200+ runs at 2.0 GHz. Try to let it run at 2.2 GHz with 1.1 V. When you're unstable, burn in again.

 

[BarboBot]

I've just started the 1.32v session (8hrs) loaded up fine idle was a little warmer than previous sessions but i think its cause i've done so many back to back or maybe the lower voltage. workin my way down to 1.1v i'm hoping after this is done i can hit 2.2 at stock voltage.

 

[rise]

you should be better than that. my winny does stock volts up to 2350 or so.

 

[garkon]

cool, more interesting posts, good luck barbo, and keep us posted on results.

 

[BarboBot]

thanks for the support i haven't seen too much info on this forum about burn in so i figured it'd be the perfect place to post results from my experiment on and get the knowledge out there. ok with that said on with the update.


All i can say is i'm extatic with the results today there was a slight temp spike with my last burn in session (1.32v was at 37c load), but this session is as the previous ones yeilding even better results i'm down to 1.29v and i'm about 1 hour into the burn in session and its at 33c under load! (now thats some improvement in my opinion, most noticable so far) I also have to add it booted into windows perfectly fine at this voltage also (its got me wondering how low i can take this burn in process i was hoping for low 1.2v to high 1.1v but now i'm thinkin i might be able to push it down to just 1.0v which would be beyond me) i also haven't encountered any reboots or crashes as of yet (haven't been running prime95 just straight burn in. I'll keep the posts coming as the sessions come and go.

 

[sandorski]

Originally posted by: garkon
Dose this work? Seems like it would not really affect stability too much, as i've never had stability problems w/ just installing OS updating all the way, running a anti vir, and keeping it pretty clean. Dose it have to do w/ overclocking? I just figured that if u were overclocking ud want to up voltage. I know nothing about overclocking, actually interested in an answer from someone who knows what there talking about!!

No. Burn In is hokus pokus. Doesn't hurt anything, so if one wants to go through the ritual there's no harm in it.

 

[BarboBot]

hmmm so i guess that makes the temp drop and voltage stability my imagination and everyone else who has reported results from their burn ins (check any overclocker forum for verification on that). It could be a case by case issue were in this case the temperature has dropped off and the voltage stability is there thus giving me reason to believe and say that going through the burn in ritual does work. Now i'm not saying you're totally wrong here, but to say its complete hokus pokus is just rediculous in my honest opinion (granted i'm halfway into this process and i've seen the results and i've seen my temps drop), but you're not the only person who has said that to me (most of my friends have said burn in is a waste of time also prompting me to try it and run this test). From the 3 people including yourself who have called it waste of time, hokus pokus, etc i've read countless articles and review that have shown otherwise (my own personal test has show otherwise) so i'm going to have to firmly disagree with you on it being total hokus pokus (p.s. you should never make a generalized statement seeing as how most of the time a "generalized statement" can't be backed up with fact)

 

[Capt Caveman]

Not sure to believe it or not but I'm' guessing a number of the times it's probably the thermal paste burning-in and setting-up properly.

 

[sandorski]

Originally posted by: Capt Caveman
Not sure to believe it or not but I'm' guessing a number of the times it's probably the thermal paste burning-in and setting-up properly.

Pretty much. Heatsink paste requires a length of time to set, not necessarily burn in.

 

[sandorski]

Originally posted by: BarboBot




hmmm so i guess that makes the temp drop and voltage stability my imagination and everyone else who has reported results from their burn ins (check any overclocker forum for verification on that). It could be a case by case issue were in this case the temperature has dropped off and the voltage stability is there thus giving me reason to believe and say that going through the burn in ritual does work. Now i'm not saying you're totally wrong here, but to say its complete hokus pokus is just rediculous in my honest opinion (granted i'm halfway into this process and i've seen the results and i've seen my temps drop), but you're not the only person who has said that to me (most of my friends have said burn in is a waste of time also prompting me to try it and run this test). From the 3 people including yourself who have called it waste of time, hokus pokus, etc i've read countless articles and review that have shown otherwise (my own personal test has show otherwise) so i'm going to have to firmly disagree with you on it being total hokus pokus (p.s. you should never make a generalized statement seeing as how most of the time a "generalized statement" can't be backed up with fact)

I'd like to see some of these "countless articles". Here's an article that's not Computer
Electronics specific, but deals with a similar train of thought popular amongst some Audio/Videophiles: Link

From that article you may notice that Electronics Manufacturers did in fact "Burn-in" their products, but it wasn't to make them work/perform better, it was to make sure the components wouldn't fail after delivery to the Consumer. That way they could avoid angry Customers and replacing flaky products later.

 

[BarboBot]

thats true but my thermal paste has been set in i've used cpu burn in (traditional method) on other rigs this is my first time testing out this method to see if it differs any my temps with this computer at stock voltage were matching those of comps with the XP-90 heatsink and the thermalpaste well worked in over a month (average 32c dle/average 35c under load) thats where i've started the testing witht his method. I'm just saying don't be so quick to judge because a. whilst i do swear by cpu burn in on most cpus and i have seen results in the post i've stated this is just a test. b. with the info you've got i'm doing this test on a base comparison (compared with average temp of my same cpu and heatsink). c. i've yet to have one person give me cold hard facts to say that cpu burn in "in general" is "hokus pokus" (granted not every application will take the same it depends on the person performing it and how much time they're actually burning in the cpu and how precise they are about it because with anything human error does come into play) d. I've read many topics on the subject with the exception of this forum (which has great usually unbiased info about a lot of other topics though not as OC inclined as most other forums i read) in which the couple of posts i've read have been pretty much passed off as the users imagination.
all in all most of what i've seen so far is a lack of fact supporting either side so i'm openly conducting this test and posting "my" real time results as i get them and it'd be much appreciated if this was taken as such (a test) and then when all is said and done and i move it back to stock voltage lets see what happeneds (give it time you could very well be right).
for a recap i'm testing to see if there any stabilitity, temp, and voltage variations after the the burn in period is up. (remember this is only a test)

 

[BarboBot]

From that article you may notice that Electronics Manufacturers did in fact "Burn-in" their products, but it wasn't to make them work/perform better, it was to make sure the components wouldn't fail after delivery to the Consumer. That way they could avoid angry Customers and replacing flaky products later.[/quote]

fair enough but like you said thats electronics and it isn't computer specific and also the first article i've read thus far.