when to restart a PC

[MTJ=DK=]

OK, that's a

1. Waste of power
2. Waste of computer life
3. I DON'T GET IT!!!!!!!!!!!

If I sit on my computer for 6 hours a day, why should I leave it on for 24? I just don't get that concept. Your computer just doesn't run as properly as when you just start it up.

+1
Only use mine 3hrs a day so way keep it turned on??

[Rand{CLR}]

I turn my off for vacations or any time I'm out of the house overnight. It was off this week for a few moments, but only because a blown light bulb (Duke...) blew the whole circuit in my Man Cave and shut everything off.

-Rand

[Duke{CLR}]

I turn my off for vacations or any time I'm out of the house overnight. It was off this week for a few moments, but only because a blown light bulb (Duke...) blew the whole circuit in my Man Cave and shut everything off.

-Rand

Was it one of those CFC light bulbs? I have never seen a regular light bulb blow a circuit.

[mapes]

DB quote

Absolutely untrue! If you talking about a harddrive, it could affect the drives life but not the computer life. If your talking about motherboards, etc, no the increased number of times you power on your equipment will in fact decrease it's life.

Yes this is technically true however in reality it is a negligible effect.

[Rand{CLR}]

Was it one of those CFC light bulbs? I have never seen a regular light bulb blow a circuit.

Nope. 3-way lamp. It was stuck on low, meaning part of the bulb had gone, and suddenly kicked up to high. A few moments later it took out the circuit. Not cool.

But I'm in a 22-year old townhouse built in 1987 when poor construction was common. My wiring is functional but we lose lights more often than we should, and my plumbing system is almost 100% that plastic stuff that was banned in the late 80s/early 90s.

-Rand

[Duke{CLR}]

Many of the homes that were built in that time period in my neck of the woods have similar issues. The worst I saw was that the siding of a garage was placed on the solid insulation which was stabled to the studs. There was no plywood or particle board used. There is nothing worse then a lame developer.

[Apache Warrior]

I remember when they switched from copper wirong, in homes, to aluminum.
Apache

[K PhaNTOM]

I remember when they switched from copper wirong, in homes, to aluminum.
Apache

I bet you also remember when they initially introduced electricity

[mapes]

I remember when they switched from copper wirong, in homes, to aluminum.
Apache

Aluminum wiring = bad!

[Apache Warrior]

Yep a lot of houses have burned down because of aluminum wiring. When I lived in Texas there were a lot of house fires attributed to aluminum wiring.
Apache

[DougBob]

DB quote

Yes this is technically true however in reality it is a negligible effect.

From my experience, which is not trivial, I've noticed it can cause anywhere from 10 to 30% degradation in solid state components. I've been working on and maintaining teletype, messaging, computer, crypto, networking and electronic devices since 1986 and I base my observations on that experience.
I'm not stating this to blow my horn or start any "my experience is broader than yours" type discussions. I just wanted to give a little on my background.

NOTE: DO NOT confuse rebooting with powering cycling(powering off and back on). There are a different sequence of events that takes place resulting in reboots being less damaging than power cycles.

I can safely say if I took two identically populated server rooms with a variety of equipment and in one I run it 24/7 and the other I consistently power cycle it at least a minimum of twice per day, you would be astounded on the failure trends that would develop in each room of hardware.
The 24/7 run server room would have break downs of PSUs, harddrives and fans, etc. (largely moving parts related components)
The consistently power cycled server room would have a significantly increased failure rate of motherboards, video cards, backplanes, etc. (largely solid state related components).

[mapes]

Interesting point however when talking about a home system and not a data center rack mount system I still say the effect is negligible over the expected life of a computer system. Rack mount pizza boxes have a higher failure rate over all from decreased air flow in my experience. Also the temp delta is huge between a fully cooled data center and a home office environment. One other point needs to made and that is manufacturers test they're products by power cycling them and running them continuously. These products have an expected lifetime MTBF. MTBF takes into account power cycling.

Some interesting reading I stumbled upon recently actually suggests that leaving the a system powered on is worse for solid state components. This is due to the effect known as Electromigration. It is especially interesting to note that over-clocking a processor and the ram bus will hasten this failure and that it is unavoidable over time. So to me thats kinda of an interesting dichotomy. It's ok to over-clock the proc/ram/mb for the trade off of power vs failure. However it's not ok to power off a system for the trade off of saving power vs failure.

Interestingly enough Intel's SpeedStep or AMD's PowerNow! which slows down the proc when not in use will actually increase the life time of the device.

From my experience, which is not trivial, I've noticed it can cause anywhere from 10 to 30% degradation in solid state components. I've been working on and maintaining teletype, messaging, computer, crypto, networking and electronic devices since 1986 and I base my observations on that experience.
I'm not stating this to blow my horn or start any "my experience is broader than yours" type discussions. I just wanted to give a little on my background.

NOTE: DO NOT confuse rebooting with powering cycling(powering off and back on). There are a different sequence of events that takes place resulting in reboots being less damaging than power cycles.

I can safely say if I took two identically populated server rooms with a variety of equipment and in one I run it 24/7 and the other I consistently power cycle it at least a minimum of twice per day, you would be astounded on the failure trends that would develop in each room of hardware.
The 24/7 run server room would have break downs of PSUs, harddrives and fans, etc. (largely moving parts related components)
The consistently power cycled server room would have a significantly increased failure rate of motherboards, video cards, backplanes, etc. (largely solid state related components).

[DougBob]

That's wild, I didn't realize they had started calling this something else also. It use to be call electron erosion.

I guess they wanted to fancy up this vocab also

Interesting point however when talking about a home system and not a data center rack mount system I still say the effect is negligible over the expected life of a computer system. Rack mount pizza boxes have a higher failure rate over all from decreased air flow in my experience. Also the temp delta is huge between a fully cooled data center and a home office environment. One other point needs to made and that is manufacturers test they're products by power cycling them and running them continuously. These products have an expected lifetime MTBF. MTBF takes into account power cycling.

Some interesting reading I stumbled upon recently actually suggests that leaving the a system powered on is worse for solid state components. This is due to the effect known as Electromigration. It is especially interesting to note that over-clocking a processor and the ram bus will hasten this failure and that it is unavoidable over time. So to me thats kinda of an interesting dichotomy. It's ok to over-clock the proc/ram/mb for the trade off of power vs failure. However it's not ok to power off a system for the trade off of saving power vs failure.

Interestingly enough Intel's SpeedStep or AMD's PowerNow! which slows down the proc when not in use will actually increase the life time of the device.

[K PhaNTOM]

OK I read all that Mapes said and still don't know if he's pro/anti power off.

[Hammy]

OK I read all that Mapes said and still don't know if he's pro/anti power off.

Yeah- me too. Sounds like this needs to be moved to the Battlegrounds.

ROFLMAO