I have heard many users and reviewers talking about a burn in period after they install a new CPU with a fresh coat of thermal paste. Is this done to cure the thermal paste and does it help at all or is it just supersticion? If it is worth it, how do I accomplish it? Should I start right out by overclocking and heating the hell out of my chip or should I gradually ramp up the heat? Thanks.
![[H]ard|Forum](/styles/hardforum/xenforo/logo_dark.png){kind=logo}
-
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.
Burn-in period?
- Thread starter MiXdNuTs
- Start date
Generally any thermal paste you buy has some kind of curing time, like arctic silver for example says in their instructions that it may take up to 200 hours of running time for the compound to fully cure and achieve "maximum particle to particle conductivity" so the burn in is partially for that, and some people say that after letting their cpu run for so long kind of "breaks it in" and will allow them to achieve a better overclock, but nothing about that is proven. Needless to say, take it to as high as you can get it stable and after a week or two see if it will go any higher.
the difference of having burned-in thermal paste is not more than like 5 degrees.
give your box a week or two to run and then try a higher OC
YMMV, but I'd rather know that I am a little lower than the limit of my equipment, it assures me that it'll run longer
give your box a week or two to run and then try a higher OC
YMMV, but I'd rather know that I am a little lower than the limit of my equipment, it assures me that it'll run longer
hrdwrjnkie
n00b
- Joined
- Jan 21, 2003
- Messages
- 59
The way that I do all my procs (not neccessarily the "correct" way, just my personal technique) is to set the voltage as high as it will go 'til I get temps in the 50C range. This is done as stock CPU speeds. I then let it run for about 48 hrs of Prime95 to heat it up real good. Two identical chips, one that has undergone the procedure and on that hasn't, usually see about a 5% or so better OC.
hrdwrjnkie
n00b
- Joined
- Jan 21, 2003
- Messages
- 59
It seems to toast up the AS3 faster, giving me better temps, and also ssem to prepare the CPU for future abuse.
accidentalsuccess
Gawd
- Joined
- Dec 7, 2001
- Messages
- 596
i had the link but lost it again.
"burning in" actually works. It effectively burns out some of the 2bizzillion transistors in the chip (the "worst" ones) allowing you to use less voltage to achieve the same speeds. It also helps w/ thermal goop, too.
I know for a FACT that my rig was not prime95 stable at speed until I set vcore at 1.775 (1.73 read volts 'cause nf7 undervolts). After 2 weeks I bumped it down to 1.750 (and whatever that is "really") and it's prime95 stable now. Weird, but it works. Saved me a few degrees, too.
It's fun to tweak out your OC twice (or more
).
"burning in" actually works. It effectively burns out some of the 2bizzillion transistors in the chip (the "worst" ones) allowing you to use less voltage to achieve the same speeds. It also helps w/ thermal goop, too.
I know for a FACT that my rig was not prime95 stable at speed until I set vcore at 1.775 (1.73 read volts 'cause nf7 undervolts). After 2 weeks I bumped it down to 1.750 (and whatever that is "really") and it's prime95 stable now. Weird, but it works. Saved me a few degrees, too.
It's fun to tweak out your OC twice (or more
). Captain Rehab
Limp Gawd
- Joined
- Feb 29, 2004
- Messages
- 279
Wow. These are some pretty fanciful speculations.
"burning" in a pc is nothing more than putting stress on your hardware for extended periods of time to identify possible faulty hardware or hardware conflicts. There are various ways of doing this, using different software.
Voltage does not fry a cpu. Heat fries a cpu.
Thermal paste does not "cure". It works to maintain heat transfer between core and heatsink or it does not.
"burning" in a pc is nothing more than putting stress on your hardware for extended periods of time to identify possible faulty hardware or hardware conflicts. There are various ways of doing this, using different software.
Voltage does not fry a cpu. Heat fries a cpu.
Thermal paste does not "cure". It works to maintain heat transfer between core and heatsink or it does not.
Capitan took the words out of my mouth.
A burn-in is something manufacturers do on motherboards for example. Its also something a system builder can do to reduce the chance of having hardware go bad a week later. And it also tests the combination of componants for flakyness.
You could also test an overclock and call it a burn-in.
A burn-in is something manufacturers do on motherboards for example. Its also something a system builder can do to reduce the chance of having hardware go bad a week later. And it also tests the combination of componants for flakyness.
You could also test an overclock and call it a burn-in.
mewannafastpc
Gawd
- Joined
- Sep 27, 2003
- Messages
- 768
voltage may not fry the cpu but it will kill it...for P4s there's a saying..."SNDS" (sudden northwood death syndrome) this happens when people on air cooling pump their voltage over 1.8vs for extended periods of time
and i am not using a northwood im using a willamette so my voltage is still in the safe side
My P4P800-d "burned" in my cpu for me. It had a faulty voltage regulator or something...anyway, it would fluctuate the voltage all the time from 1.50v all the way up to 1.87v...with bios set to 1.52v.
Had a brand new 3.0C in there running stock and after a week it wasn't even stable at 3.0. dropped it to 2.8 and still had some random problems...and this was in two motherboards.
RMA'd both of them.
So if you ask me if voltage can kill a cpu...my answer would be YES. enough voltage will kill anything. Electronics don't survive high voltage just like they don't survive heat. You can supercool a cpu and feed it 1.9v and it will still die.
Had a brand new 3.0C in there running stock and after a week it wasn't even stable at 3.0. dropped it to 2.8 and still had some random problems...and this was in two motherboards.
RMA'd both of them.
So if you ask me if voltage can kill a cpu...my answer would be YES. enough voltage will kill anything. Electronics don't survive high voltage just like they don't survive heat. You can supercool a cpu and feed it 1.9v and it will still die.
allenwr1505
Weaksauce
- Joined
- Oct 16, 2003
- Messages
- 106
Does a burn-in period work for memory? I know that it worked for my processor, because I was able to run it at higher speeds after a few weeks. Most likely that is just because of the AS5 curing though.
Captain Rehab
Limp Gawd
- Joined
- Feb 29, 2004
- Messages
- 279
And so what happens? I suspect that the temps get too high and damage the core. But, not having any personal experience with northwood chips I won't deny that it is possible. I'm just not [H] enough yet
lol
let me put it easy
voltage only kills p4s, unless you set an AMD's voltage to something over 2v.
artic silver DOES NEED TO BE CURED. temps DO lower after about 200 hours running time. Making it hotter will NOT improve this.
IN MY EXPERIANCE, CPUS DO GET BETTER OCS IF BURNT IN. When I first got my CPU I couldnt run 2.5 @ 1.825 volts without getting prime 95 errors. a month at 2.4@ 1.725 and now I get NO prime95 errors at 2.5.
voltage only kills p4s, unless you set an AMD's voltage to something over 2v.
artic silver DOES NEED TO BE CURED. temps DO lower after about 200 hours running time. Making it hotter will NOT improve this.
IN MY EXPERIANCE, CPUS DO GET BETTER OCS IF BURNT IN. When I first got my CPU I couldnt run 2.5 @ 1.825 volts without getting prime 95 errors. a month at 2.4@ 1.725 and now I get NO prime95 errors at 2.5.