How many ppl burn in their cpus?

[Anonemous]

I was reading around and I noticed therewere some ppl who suggested just ocing the cpu right out of the box while others suggested a slow burn (24hours+) in of the cpu @ high voltages (1.9V). I was just wondering if burning in the cpu @ high voltages at a low multiplier and the highest stable fsb has any effect on the overclocking ability of the cpu.

 

[SinfulWeeper]

I think it's a myth. I never do it...

But thats only me. Others will agree/disagree.

 

[DAPUNISHER]

Originally posted by: Anonemous
I was reading around and I noticed therewere some ppl who suggested just ocing the cpu right out of the box while others suggested a slow burn (24hours+) in of the cpu @ high voltages (1.9V). I was just wondering if burning in the cpu @ high voltages at a low multiplier and the highest stable fsb has any effect on the overclocking ability of the cpu.

read the Anandtech FAQs before posting questions SinfulWeeper is correct and it doesn't really matter to me who argues because this info comes from highly knowledgable individuals and pm works for intel so I think he may have a better idea than most on the subject He's mentioned in a couple threads that they stress test CPUs in in very expensive proprietary equipment designed soley for the purpose that put loads and stress on the CPU that we the end-user could not hope to place on them. Therefore when you recieve your CPU it's ready to rock&roll.

 

[amcdonald]

I don't burn-in chips. My dad used to engineer microprocessors, and when I asked him about it he didn't understand why people would ever do this.

 

[pspada]

P.T Barnum was right - there IS a sucker born every minute. I need to buy a new bike, should I put it in the oven for a few hours at a high tempeture, because it'll then go faster? How about a pair of Nikes? Think a few hours at high temp will make them run faster too?

 

[DAPUNISHER]

Originally posted by: pspada
P.T Barnum was right - there IS a sucker born every minute. I need to buy a new bike, should I put it in the oven for a few hours at a high tempeture, because it'll then go faster?

I'm a mountain Biker so I found that humorously ironic I ride a Cannondale I've had for years with a full suspension 6061 series aluminum frame and they do a seamless TIG weld on it then stick it in the oven because it redistributes the aluminum's strength after the welding weakens it around where the weld was performed. It rsults in a stiffer and stronger frame and stiffness actually does to a very minute degree improve speed by transfering more of the energy to the pedals instead of having the frame absorb some. Of course I ride full suspension so it's not much help but with a hard tail it is.

 

[tdowning]

well, I do burn-in, but it's a much different process for me, and I only do it where stability is absolutely crucial, and, come to think about it, I've only done it on win 98 rigs.

I will usually only install one driver every 24 hours or so, (leaving the computer on, and running a Distributed Computing program to max out the CPU) so that I know which driver (If any) might make the computer unstable. (Or if there is a problem before I install any drivers, I look into flashing the BIOS)

I think I've quit doing it because most hardware comes with XP WHQL drivers, and you can bet MS spends lots of time testing those, and I don't think I've ever had any major stability problems on XP.

 

[Corporate Thug]

no burn in here, i think its BS

 

[screw3d]

Agreed.. My 1700+ ran 2300MHz right out of the bat LOL.

 

[BOBDN]

I don't burn in but I stress test the system for stability. Sandra burn in wizard X 10 and Prime 95 12 to 24 hours depending on how much time I have.

 

[Insidious]

Many ICs are burnt-in as part of the QA process. (Identifies/accelerates infant failures to avoid public exposure, etc.... )

Don't they do that with CPUs as well?



Would I consider intentionally over stressing electronic components.....? Not me baby!

-Sid

 

[amoralist]

system burnin isn't for speed. the idea is that electronics (as opposed to mechanical devices) generally will fail (if they will fail at all) within the first 48 hours... so that's why the rule for system builders is to "burn in" or run the comp for 48 hours before giving it to their customer. that way it won't die on the customer.