Northwood P4's dying prematurely?

[Scyphe]

Have you guys read the news about NW P4's dying prematurely because of overclocking? I just saw an article at The Inquirer (yeah, I know, tech-tabloid) with links to Aces Hardware where they keep a database of P4's that died of overclocking.

http://www.theinquirer.net/?article=6549

Makes you wonder if that's a new Intel strategy to force their customers upgrade instead of overclock and at the same time make a buck when peeps try to overclock (weee, talk about paranoia) :rolleyes:

[FUGGER]

I guess you do not drive your car because there is a chance of getting in a accident, you do not fly in planes because they crash, you do not go out side because you will get struck by lightning.

You have a better chance at one of the above than killing a P4 CPU.

[Colin]

If you kill a chip, just RMA it.

When I killed my old 1.6A, I just told them that it stopped working and that I was a newbie. :)

[Scyphe]

Hey guys, don't kill the messenger.. ;)

I only found it curious that certain Northwoods goes down due to overclocking. But as we all know, The Inquirer is really a tabloid site with news to take with a grain of salt. But if there's any truth to it, it sure is weird.

I don't even use Intel anymore even though they're the current king of overclockers, too expensive with "better bang for the buck" in mind when I upgraded to AMD last time. Besides, I sometime go with the underdog instead of the big bad wolf. ;)

[SupaMan]

Hmm, interesting. sux to be them :)

AMD!!!!! :D

[NoEcho]

Interesting. I've heard that the tiny gaps wear out quickly (within a few months) with heavy overvolting. Absent that the only risk seems to be heat... and with throttling that risk is of slowing down, not burning up.

My mobo has a limiter: 15% overvolt max. It won't LET me push it into the orange zone....

Unless 15% IS the orange zone. However, since that's just a .175 V increase I doubt it.

Most interesting thing I saw in that thread was the premise that maybe at .13 micron we've reached the point where even regular use will degrade the gaps... that maybe all cpus have become disposable.

[wwilliam54]

I think it from the p4 haveing too few metal interconnect layers
5 or 6
compared to the tbred b's 9

[majormav]

to be honest one thing about the retail northwoods is a 3year warranty if its dying just kill it and get a new one simple as

[SupaMan]

Originally posted by majormav
to be honest one thing about the retail northwoods is a 3year warranty if its dying just kill it and get a new one simple as

bingo

[Wolf]

Damn that sucks for Intel guys. I've been running 2.3v through my .18 pali for a year with absolutely no issues. I think it has something to do with the number of layers Intel uses on the chip compared to AMD's and the .13 die size. I noticed this trend this summer and it's not looking good unless you buy a retail chip. I am confident AMD's new chip will hold up much better under voltages than these fragile p4's have. All I expect is to get a year out of a cpu but these chips look like they are failing in less than 6 months with over voltage. :eek:

[KnightElite]

The reason that P4's "die young" as it were, from overclocking, has to do with the way that MOSFETs (the transistors used in digital logic circuits) work. Basically, a voltage is applied to a gate, which is seperated from the channel through which the current flows by a silicon dioxide (non conductive) layer. That voltage controls whether the transistor is on or off. However, when too much voltage is applied, the electrons may bore through the insulating oxide layer, which is quite thin, and short the gate to main transistor channel. That kills the transistor, and if it only happens once, it's not a big deal (a few random errors maybe) but if it happens lots, I would imagine it would completely screw your CPU over. This is called Gate Oxide Breakdown. The reason that this happens is due to Intel's design on the 0.13u P4s. So far, it doesn't seem to happen on the .13u Thoroughbreds, but apparently AMD used a .13u gate even in the .18u chips in any case, so we don't have to worry about this particular issue on Athlons, at least not yet.

Hopefully that made sense, since it's 4:30 AM, and I haven't got to bed yet ;).