Will CPU burn in work? Whats the best program?

[Mellow_Dude]

Hi Guys,

I finally got my PII300 SL2YK working at 450Mhz (Voltage up at 2.18v by taping pins).

I hope to drop back to 2v if I can once the CPU has run in at 450Mhz. Will this work? Do I need to run any apps to burn in the cpu? Is their any truth in this theory?

Many thanks,

Mellow Dude

[Alegria]

Normally the Burn-In-Process is used on a variety of electronic components to insure the reliability of the parts under certain circumstances such as a given temperature, humidity and so on. It is mainly used to age components artificially in a shorter time than possible (if it still works after the Burn-In, it's supposed to go perform for the estimated life-cycle of the part). As far as Microprocessors are concerned there is no such thing as overclocking at a higher CPU-Core-Voltage and then dropping back to the default voltage just to see that it still performs reliable then at a higher clock speed. It works for you at 2.18V. Does it work at 2.0V? If it does I would leave it alone immediately and insure proper cooling.
If it does work at 2.18V, fine, still insure proper cooling and leave it like this.
I won't go into the technical issues of degradation of aluminium leads and so on but it is true that the lifetime of your processor will be shortened by overclocking.
Who cares though? You probably might want to change your CPU long before the side-effects like unpredictable crashes of your machine kick in. Given the higher clock frequency on the frontside bus not only means your processor performs faster but the on-chip cache is put to the limit. If you want to find out if it performs well, try this one:

CPU Stability Test
get it at http://www.saunalahti.fi/jv16/

Regards, Sascha Alegria Mayer

[Greebe]

Alegria, Thank you!!!

------------------
Designer of the first commercial Athlon mod. Currently marketing the "Eclipse" through AZZO Computers.
Anyone up for an Eclipse'd Athlon 850? Email me for more information

2836 mips, 1.107 gflops @ 850 mHz 1:3 latency
CPUmark99 72.3
FPU Winmark 4630

3Dmark 99 Max (8x6 16, 16, triple) PD5.41
7885
15181

[Nuno]

before the side-effects like unpredictable crashes of your machine kick in

It´s worth noting that most of those long-time unpredictable crashes are cause by a poor quality thermal paste that dries up and the cpu just over-heat.

[aNtHrAx323]

whoa, hey now. CPUs can be burned in. i have a K6-III 400, and when i first tried to run it at 480 (4.0 x 120), it would only POST at 2.9V and up(a helluva lot of heat... 130º F w/ a peltier and an Alpha). but, after running quake 2 on it nonstop for about 48 hours, i lowered the voltage to 2.8, then 2.7, then 2.6... all the way down to 2.3V. now it is perfectly stable at 2.3V, and at sub-85º F temperatures. so yes, CPUs can be burned in.

------------------
-Campbell "aNtHrAx323" Krueger
anthrax32@yahoo.com

System specs:
AMD K6-III 400 at 112x4.0 at 2.3V
Asus P5A
96 MB SD-RAM (the old 66 MHz... stable enough tho =)
Matrox G400 AGP 32MB Dual-Head
Diamond Sonic Impact S90 PCI
Linksys LNE100TX WOL PCI Ethernet adapter
Processor is a peltier/Alpha cooling combo, avg. temp is around 85 F.

http://anthrax.home.texas.net

[Wombat]

Yes. Burn-in works. My system is proof. I agree completely about OC'ing reducing product lifetime, but you're mistaken about burn-in. I study the subject, and it seems like burn-in should weaken the CPU, but it doesn't.

[Greebe]

Believe what you wish, but I've never experienced any situation that proved that a "burnin" process enabled ANY cpu to run faster. This trend came about from most likely an overzelious electronics tech that figured the burnin that's done to "green" crystals would apply to cpu's also. This is incorrect. CPU's do not frequency settle from such a process. If you make the claim that you could now run at a lower voltage after cooling it with a peltier device then that proves my point. See when cpu's are cooled to a low temperature it's similar to raising the voltage, but when cooling like so they don't tend like the higher voltages. If your wondering if this is just me tooting my horn then ask Gabriel Rouchon of Swiftech. He'll backup what I'm stating here also.

------------------
Designer of the first commercial Athlon mod. Currently marketing the "Eclipse" through AZZO Computers.
Anyone up for an Eclipse'd Athlon 850? Email me (epc@netdirect.net) for more information

2836 mips, 1.107 gflops @ 850 mHz 1:3 latency
CPUmark99 72.3
FPU Wi

nmark 4630

3Dmark 99 Max (8x6 16, 16, triple) PD5.41
7885
15181

[Alegria]

As always some agree and some don't, therefore I would like to mention that I'm not some homemade expert. Please read an excerpt from the Military standards about burn-in testing:
Now what is the purpose of BURN-IN-TESTING?

MIL-STD-883E
METHOD 1015.9
BURN-IN TEST

1. PURPOSE. The burn-in test is performed for the purpose of screening marginal devices, those with inherent defect or defects from manufacturing which cause time and stress dependent failures. In the absence of burn-in, these defective devices would be expected to result in infant mortality or early lifetime failures under use conditions. Therefore, it is the intent of this screen at or above maximum rated operating conditions or to apply equivalent screening conditions which will reveal time and stress dependent failure modes with equal or greater sensitivity.

There you go.
Yours, Sascha Alegria Mayer, who studied the
subject.

[Greebe]

I too have studied the subject and practiced it too. The purpose of that specific mil spec is only to insure that early failure won't happen on mission critical equipment and does not lend itself to extending or improving the quality or proformance of the part.

[Plmr]

Like the bumble bee that defies the laws of aerodynamics my p3 450 topped out at 558 when I got it, now after 3 months 1t is stable at 620.No bios ,cooling or other hardware changes execpt speakers.Hmmmmmmmmm Curiouser And Curiouser... Matt

[Alegria]

I forgot to give an interpretation to this mil spec. Of course it is not the intention to extend the lifetime of a specific part of equipment but for quality assurance purposes. I chose this as an example for burn-in testing because it explains it as it should be. Whether or not being a mission critical part or machinery, testing is done to insure proper function.
Overclocking beyond specifications always bears the risk of malfunction or failure of microprocessors. If something works overclocked, fine with me, if not, what the hell...
Does anybody dare to ask Intel, AMD or DEC/Compaq/Samsung about overclocking Pentiums, Athlons or Alphas? Let me know their answers on the subject of burn-in testing of their processors...Surely won't be "Go ahead, burn-in beyond our specification provides you with a higher rated processor than we could find out using sophisticated testing equipment." Of course there have been underrated processors because of better production procedures. At the moment AMD would be glad to reach a higher yield rate with their top-of the line processors. But many of the CPUs don't perform properly during testing. So they are rated lower than intended. Thus they are cheaper than higher-rated CPUs which passed testing. My bottom line to all this is:
Do whatever you wish and believe or trust whoever or whatever you think is right.
It doesn't matter anyway.

[Plmr]

I guess I lied about "No Bios Changes"

Matt