1

According to Energystar.gov, modern PCs are built to handle 40,000 on-off cycles before failure. Does on-off cycle frequency make much difference to component wear?

Consider 3 sets of identical laptop computers:

Control: 30 on-off cycles every month for 10 months (standard consumer usage)

Sample 1: 30 on-off cycles every day for 10 days (heavy testing)

Sample 2: 30 on-off cycles every hour for 10 hours (excessive testing)

Criteria: Power-on duration per cycle standardized in all three groups to 1 minute. Power supplies remain connected and plugged in (active) when laptops are powered down.

Would there be a noticeable difference in wear between these sets of computers?

2
  • This question may receive a great deal of speculation and anecdotes for Answers. I was about to give an anecdote myself, then realized it's not a good Answer. Sep 5, 2018 at 22:23
  • If you have a substantive or interesting anecdote, please share it. It may turn out to be helpful in some way. Sep 5, 2018 at 23:02

1 Answer 1

1

Each of the tests consume 300/40,000 (0.75%) of the life of a modern PC, accepting the hypothesis of the original poster they quoted from that .gov site. The wear on each at the end of testing is the same, since they each undergo the same number of test cycles, just over differing timeframes.

Now, if you want to talk about iatrogenic wear compared to remaining life....

The forecast life of that laptop PC before it either fails or is no longer useful because of code bloat and/or rising expectations is quoted as 3-5 years. Let's make that four.

Control PC: At the end of 10 months, you've got 38 months to go to reach end of life, or 79.2% of the nominal utility, but only expended 0.75% of its nominal lifespan. ANSWER: NO.

Sample 1 (heavy): At the end of 10 days, you've got 1,451 days to go (99.3%), but expended 0.75% of its nominal lifespan. The pendulum is starting to swing, so ANSWER: MAYBE.

Sample 2 (excessive): At the end of 10 hours, you've got 35,054 hours to go (99.97%), but expended 0.75% of its nominal lifespan. ANSWER: YES

but a laptop PC by then will come free in cereal boxes.

4
  • 1
    Thanks for the reply, and for updating it accordingly. Do you have a source or reference for the first paragraph of your comment? Sep 5, 2018 at 22:42
  • 1
    I'm presuming the energystar figure to be based on standard consumer usage. The question is whether 'excessive' or 'heavy' usage beyond standard usage increases component wear and reduces the maximum number of cycles before failure. Sep 5, 2018 at 22:54
  • 2
    I was told in 1988 by an ITT semiconductor engineer the major cause of circuit failure is cycling power back on within less than an eight second interval. The 'excessive' regimen hypothesized here allows for a two minute cycle, much longer than that eight seconds which leads to damage.
    – K7AAY
    Sep 5, 2018 at 23:00
  • 1
    That's a very interesting anecdote. Thanks. It might be worth investigating this to see if there is at least some additional protective effect of powering off longer than 8 seconds. Sep 5, 2018 at 23:12

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