Super User is a question and answer site for computer enthusiasts and power users. Join them; it only takes a minute:

Sign up
Here's how it works:
  1. Anybody can ask a question
  2. Anybody can answer
  3. The best answers are voted up and rise to the top

I recently had a motherboard fail on me. If I add my experience with motherboards of desktops, laptops and servers, I've probably been responsible for about 50-60 years of motherboard use (most in a desktop environment). The failure I experienced was in a server that ran 24/7 for the past 24 months (new machine in 2008). Now I've been in the biz for about 20 years, so some of the 'long life' motherboards were using 1990s technology.

Can anyone comment on whether a 50 year mean time between failures (MTBF) is fairly typical for all motherboards that are not DOA? Especially, given 2010 manufacturing/quality techniques. Assume that the motherboards operate in typical humidity ranges in office environments that typically don't exceed 85 degrees F, and are not subjected to high shock or dust conditions, and have proper AC current conditioning and over-voltage protection. What should be expected for consumer grade motherboards for MTBF?

Incidentally, I have a server that is 2001 vintage -- I should be good for another 20+ years, right? Plus, then I can sell at auction for $10,000 in 2030 ;)

share|improve this question
up vote 6 down vote accepted

Capacitors. It's all about the capacitors.

They seem to be the first thing to go on motherboards, and an old psu with a 12v rail that flickers between 11 and 12.5v isn't going to live 50 years. If the motherobard has a good quality psu to go with it, and is kept at a steady cool temperature (so the electrolyte in the capacitors doesn't crystalise) the board will last to its full potential.

Also, am I right in thinking your MTBF calculation is likely to be skewed upwards if you're basing it off the use of multiple motherboards for short periods?

No single motherboard in your tests will reach its expected end-of-life, so you're only examining the left side of the failure rate "bell curve". To paraphrase, given you're ruling out DOA cases, running one motherboard for 50 years is going to be more likely to fail than running 50 motherboards for 1 year.

I would expect a MTBF of ~10 years.

share|improve this answer
Totally agree on the "It's all about capacitors". At least three of my past motherboards died from capacitor failure... – celebdor Nov 5 '10 at 23:41
Now that I think about it, I might be pushing my luck with the 10 year old PC repurposed as a server. Doubly risky, is running with a 10 year old hard drive. How many bullets are you supposed to use when playing Russian roulette ? ;) – Rolnik Nov 6 '10 at 4:47
Failed capacitors, cracks due to heating and cooling, failure due to poor quality PSUs, electrical storms, and power fluctuations (surges & brownouts). I've probably built and rebuilt about 50 machines over the past 10 years, and I've found that motherboards typically fail for those reasons--mostly due to poor quality capacitors or external electrical issues. – Force Flow Nov 6 '10 at 5:39

I haven't any specific numbers, but in general, the mean lifetime is well above the expected use. Considering computers are obsolete within just a few years, you can generally expect 10 years of life from some of the still parts. Hard drives, fans and power supplies have a somewhat shorter lifetime, but motherboards and in some cases memory can last quite a while.

20 years sounds a bit long to me, I personally haven't seen components running for that amount of time, but it might be possible if you get solid hardware and keep it in good conditions. I would suggest somewhere between 8-12 as the average maximum, though.

share|improve this answer

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .