new pc, burn in?

any suggestions on ways to burnin, or software to use? do i need to focus on any perticular part? i.e. cpu, gpu, etc?

 

I got an lg 16x/48X with 4mbs of cache and have always used nero its a preety good app also about the cpu i think you could get away with even a celeron 2.00 or higher but it will kind of slow down ur system so to be sure go with a pentium 4 2.5 or higher

 

oh,, my bad. i meant a burn in test. to break in new components. sorry. but thanks sumting77

 

Not really...install your apps and start using it lol

 

This aint a car, where the engine is brand new and pretty much never turned on before the buyer. A laptop goes thru a lot of burn in time while the manufacturer sets up the OS, and installs all the crap that come with new systems. How long do u think a computer will be on and running to install all those trial software etc.

Unlike a car engine (which many pple recommend a burn in), a computer cannot be sold to a customer w/o ever being turned on, unless ur gettin a barebone with no OS. So most computers do have plenty of "ON" time from the moment their are assembled to the moment u turn it on for the first time urself.

 

there is an application called CPU burn-in that is designed just for what you ask.. you gotta run it twice at once though for dual core.

 

"Burn-in"s are mythical procedures. Generally, they stemmed from checking out if an overclock succeeded by running something "horsepower intensive" and checking for mathematical or logical failures (ie. Prime, SuperPI, etc.)

However, electronic components don't need burn-ins. It does not get quicker, more efficient (unlike an engine that has moving parts with physical tolerances), or anything the like.

If you overclock, you test. If you don't, you just use. If something fails, it's not because it wasn't "burned in" -- it failed simply because it was the weak chain in the link regardless.

 

oh......... okie dokie. thanks!

 

I actually have another side to this story...

New systems go through a burn-in period usually to let cooling paste set as well as to keep a consistent flow of voltage to the system to verify they work at maximum specification. Electronics are normally built using hardware that is 5-20% tolerant of their specification and the burn-in usually helps to verify that you don't have a bad mix of the worse end of the parts. Generally lot's are tested and mixed correctly to help with this, but some bad lots still get through. Burn-in's also help to make sure heat is dissipated correctly, as well as EMI isn't charging up from any improper contacts or grounding. Burn-in's for notebooks also generally include the first battery cycle.

As for the people saying you can't buy computers without a burn-in.. you better check again, especially when concerning white-books that are assembled by resellers, or the value seekers that buy the whitebook themselves. I know of a few resellers that just place a standard image on an HD from a cloner, slap the drive into a freshly opened never powered system, then ship it out the door... I've also worked and quit several places that do that..

 

I definitely appreciate the burn-in that Powernotebooks did on my HEL80 before sending it out to me. Nothing is more irritating than having to send your new toy back to the seller due to a faulty part.

 

Then we're mincing the terms "burn-in" with "new system diagnostics".

This is not something the regular end-user can determine, unless you're talking about holding your hand over an exhaust vent. I suppose they could have access to a thermal imager ($$$), but without knowing the exact heat dissipation, circulation, and tolerance levels of various locations within the computer, it's generally useless. If you've ever seen somebody freak out over the temperature around MOSFETs, you'll know what I'm talking about :O. I am unsure, however, how electromagnetic interference could "build up" from improper contacts or grounding. Are you referring to a static charge on the object?

So, really, these "new system diagnostics" determine if a component will fail right off the line after their recent installation or handling. However, as we all know with electronics, components will fail regardless -- it's just a matter of time. So while you may be snug in the bed glowing in the passing grade of a 36-48 hour diagnostic session that stressed CPU, GPU, RAM, drive, and optical components.. it could fail tomorrow.

Or if you have my luck, just after you complete a project you've been working on for 6 months.

Businesses will generally do these diagnostic sessions before they send you the unit because in the end, it will cut down on RMA shipping costs, technical support costs, and diagnostic costs later on -- not to mention bad reviews from shipping faulty units. It's commendable, to an extent. However, if you do have an RMA issue, lots of companies (retail, TIER 1, etc) will just ship you the replacement component without doing another diagnostic session. This essentially kills the purpose behind of the first session because this new component could fry everything upon installation and boot. Who knows.

Electronics will die. Diagnostics give you piece of mind that they won't die right now...

However, if you still prefer to do them, run these simultaneously:

1x SuperPI for each CPU core
3dmark06
HDTach
Nero DriveSpeed for optical
bang on the keyboard

continuously for 24-36 hours.

 

Of course there's confusion. Computers are just like cars, and cars are complicated! (irony )

I think I've got my head screwed on fairly straight with this.. burn in is about thermal grease, and "burn in" is about testing to make sure your overclocks don't explode the laptop. (and then "vendor burn in", which may or may not involve overclocks or explosions) *thumbs up*

So, do CPUs actually wear out? It's a little strange to think of something not wearing out, but then, it's not physically rubbing on anything... And I know those solid state storage things have a limited number of writes...

 

Yes, CPUs wear out. It's through numerous processes, one of which is called electromigration. The CPU will be replaced far before this becomes an issue, unless it was given an advanced head start due to an issue with quality control at the manufacturer.

Electromigration is an issue inherent in all current circuits. Overclocking and heat further accelerates this problem, and it's not linear.

 

I'm not confusing burn-in with anything... I was doing those back in the 486 days before anyone overclocked at all. That was the official term then and I'm sure I can dig up some PC diagnostic books published from that time that use the word the way I'm refering. Infact, I was doing those same tests under the term burn-in when the kryotech chassis came out with refrigerations and all the geeks were buying 450mhz slot A's to run them at 1.0ghz....

"Burn in is that process by which components of a system are exercised prior to being placed in service (and often, prior to the system being completely assembled from those components). The intention is to detect those particular components that would fail as a result of infant mortality, that is, during the initial, high-failure rate portion of the bathtub curve of component reliability. If the burn in period is made sufficiently long (and, perhaps, artificially stressful), the system can then be trusted to be mostly free of further early failures once the burn in process is complete.

A precondition for a successful burn in is a bathtub-like failure rate, that is, there are noticeable early failures with a decreasing failure rate following that period. By stressing all devices for a certain burn in time the devices with the highest failure rate fail first and can be taken out of the cohort. The devices that survive the stress have a later position in the bathtub curve (with an appropriately-lower ongoing failure rate).

Thus by applying a burn in, early in-use system failures can be avoided at the expense (tradeoff) of a reduced yield caused by the burn-in process.

When the equivalent life time of the stress is extended into the increasing part bathtube-like failure rate, the positive effect of the burn in is inverted. In a mature production it is not easy to determine whether there is a decreasing failure rate. To determine the failure time distribution for a very low percentage of the production, one would have to destroy a very large number of devices.

When possible, it is better to eliminate the root cause of early failures than doing a burn in. Because of this, a process that initially uses burn in may eventually phase it out as the various root causes for failures are identified and eliminated.

For electronic components, burn in is frequently conducted at elevated temperature and perhaps elevated voltage. This process may also be called heat soaking. The components may be under continuous test or simply tested at the end of the burn in period." - Wikipedia (burn-in)

I'm going to have to just shrug and agree to disagree on this one... I've done system burn in's both for intel and for many small computer shops... The thermals working correctly I'm refering to are to make sure the thermal paste is "baked", verify the system operates withing a certain temperature range for a prolonged amount of high use, AKA most parts that are faulty will show it within the first 24-48 hours (lots of blue screens, resets and whatnot can come from faulty parts only when the system is pushed to the limit for a duration of time).

EMI can bleed into the system shorting out parts.. this isn't all that uncommon from HD's or CD-ROMs... you can also get EMI from the fan motors to bleed back into the board killing the headers or the part itself. HD's and CD-roms also have a high rate of bleeding into the bridge chipsets which can cause audio distortions or visual distortions on the screen. Touchpads can also cause this EMI problem...

None of these tests require anything more then time and attention to detail... no fancy equipment is used at this point, that should only happen to select parts from lot's straight from manufacturing to verify the capabilities of the lot.

Anyways.. I'm done on this topic...

 

While I've never been one to use this term given it's ambiguous nature, we're generally saying the same thing.

I didn't think the 450s were that great.. surely not as good as the Celeron 300As back then. I missed them by about 6 months -- had to settle with a 400. Beat my old K5 and 8086, however. Though I had that 8086 in pieces more than those other two combined. Anyway..

While I'm quite familiar with EMI introducing line noise and partial interruption on a signal in extreme cases, I've yet to come across a situation where RFI created by a computer component was actively able to destroy another. This is mainly because the amount of electromagnetic radiation produced by computer components is strictly governed by international standards, and by those same standards, they're also required to function when exposed to a predetermined amount of radiation (which should be less than the maximum amount emittable). Regardless of that, EMI doesn't have much bearing on "burn-in"s, and is certainly not something the average user can account failure to.

Yes, doing those extended "diagnostics" when you first get or build a system can provide beneficial results if you're more willing to deal with RMA issues now rather than later. There's really nothing hurting somebody from carrying them out overnight or when they're away from the system -- especially on whitebook and custom systems. It can bring to light failures now under stress, rather than later when you put that system under the same amount of stress. However, I still hold that there's still nothing keeping the system from failing down the road due to any number of reasons.

Just like somebody coming across a failed CPU. These CPUs go through general "burn-in"s when still at the manufacturer (part of the speedbinning process), and they passed them to get to the shelves. However, after it got to the buyer, it croaked. So one process passed, another failed. Video cards, sound cards -- we've probably all done the defective returns line dance and they all "passed" company quality control. How many grains of salt it takes to balance that scale determines on the manufacturer.

I think there's enough information in this thread to allow somebody to make an educated guess for themselves what they think their best course of action is, whatever it may be.

 

You are missing the point here though.. You are correct in the individual component testing being done pre-packaged for resale, but you aren't including a "combined component" test that can show early signs of incompatibilities between components and other "multi-segment" faults that occur. Often drivers can aide with these issues, but not always... I know of many sound card / motherboard combo's that have issues with interrupts because the manufacturers have crap designs. I've also seen memory that works on 90% of the motherboards available, but not the other 10%. These kind of situations are really where these burn-in's are valuable.

While this is generally less necessary on systems that are rolled off assembly lines, consumer built systems aren't so simple.. And for the sake of arguement the term "burn-in" was in question... which is what I am defending...

Gotta give you kudo's for maturely defending your statement as well as provide good information to drive this thread.

***NOTE***
Celeron 300A's were kick ass for their day, but the refrigerated chassis (new technology at the time for mid-tower only worked on the socket A's... Cracked the cores on the Celeron's since the chips where exposed... I had a dual slot Tyan MB with the 300A's running at 550mhz... SMOKIN!

 

pyro ftw

 

Actually, recent research by Google shows that the bathtub curve pyro describes is not applicable to at least hard drives, and likely many other bits of hardware

http://hardware.slashdot.org/article.pl?sid=07/02/18/0420247

 

If that's what the person is after, then yes, new system diagnostics and "stress testing" can aid in determining hardware or software incompatibilities early on. They are, of course, not limited solely to that testing session and can pop up regardless down the road -- especially when exposed to something the original diagnostic did not expose the hardware or software to.

However, incompatibilities in either hardware and software (ie. drivers for those old Creative Live! cards to name one) is aside from the concern about the cooling paste, heat dissipation, voltage regulation, etc., named earlier in this thread. A "burn-in" is what I, electronics colleagues of mine, and to some extent, Wikipedia, would call a session in which a component is stressed to it's maximum rated tolerance or above for an controlled period of time to test for infant mortality -- not something Johnny EndUser can do. Somebody running prime95 to get 100% CPU usage is not really getting 100% CPU usage. Besides, better single component stress testing has already done by the manufacturer.

So, by your earlier definition of "burn-in", that doesn't really apply to compatibility testing of component combinations after their installation. We're now more into running benchmarks or various games and mathematical programs to see if you get a rounding error of less than 0.5, BSOD, or other anomalies that should not be present. A diagnostic phase, if you will. It has its own merit, I agree. Weed out early issues now while the car is still warmed up. However, it will not show everything, and issues with hardware will present themselves pretty quickly regardless of diagnostics or not. They'll probably show up within the first 5 minutes of a Windows boot -- especially IRQ sharing or memory problems. Or your first gaming session.. <shudder> Almost as tense as the first boot.

I think this thread has turned more towards a discussion about semantics. An enlightening and entertaining discussion nonetheless. I give you kudos as well, and say we end this friendly debate of terminology here and both agree to answer the original question with

"Couldn't hurt."

 

I'll accept that

 

ha. i like this forum. i'll just install the components and when they die, they die. dont have any more money after buying all of this stuff to put stress on them and possibly lose a component right now. waiting on my ram to come in (over a week now) and maybe i will finally get to turn it on and see what all the fuss is over the HEL80.


thanks everyone for the enlighting, and amusing debate.
peAce!

 

Enjoy your new notebook. And please continue to visit us and post about your experiences with your HEL80.