Bad to leave computer running all the time?

[Constant Gardener]

As for the energy savings, a PC at rest equals an average lightbulb or two.

The average light bulb in this house is a 13-watt CFL. The PCs are hungrier.

[schematic2]

I look after over 200 Pc's on site. Some are left on, some are turned off. The biggest failures I encounter are worn fans which eventually seize from use. Fans have bearings and bushings. "Running Time" is their enemy. Once the fans fail the electronics overheats.
The second biggest failure is caused by dust and dander accumulated in the processor heat sink, especialy if the tower sits on the floor. Next comes hard drives. The electronic boards and components are the least troublesome.
LS is right about the mechanics...

All B.S. aside, most computers are near outdated before they expire for whatever reason....but if you want to squeeze years of life, turn it off when not used for more than a day or two....

[handyguy]

In the old days the power supplies couldn't take the on/off cycling. But these days we have much better ones so that's not an issue now.

[Video Head]

Could the heating and cooling of cycling the power on and off not cause solders and board level components to crack and seperate? I know in mechanical applications it is ideal to have a constant temprature...cold starts with quick run-up's cause fatigue and wear, much like a car engine in winter...

[lordsmurf]

Could the heating and cooling of cycling the power on and off not cause solders and board level components to crack and seperate? I know in mechanical applications it is ideal to have a constant temprature...cold starts with quick run-up's cause fatigue and wear, much like a car engine in winter...

No more than a car. The parts are rated for ups and downs -- expected, actually. It assumes a curve of heat-up and cool-down times. It doesn't go from 70-100 in a second, it takes time that is allowed by the parts in use.

My car is fine in winter, unless it's 10-F (at which point I have to take a hair dryer outside because the stupid doors freeze shut!)

[slacker]

I think schematic2 put the practical importance of the question into perspective. Computers have gotten to the point where the physical hardware outlasts the OS and software upgrade cycles which ultimately demand more powerful systems. I often upgrade to a new computer out of software necessity, and only occasionally because something (cpu, memory, etc) has died.

[sanlyn]

It's 6 of 1 and a half-dozen of the other. Constant rebooting does create stress on the average PC. As for servers, they are pricey stuff that is "designed" to run 24/7; the average PC is not so well equipped.

It depends on how your PC is used. If it's occasional stuff like a few emails or light 'net shopping, you may as well leave it on all the time and pay the electric bill. If it's heavy graphics, design, gaming, or video work, or if you're on the 'net all day, software memory leaks will slow your system to a crawl after a few days or even a few hours. The cure is not a warm re-boot -- the cure is to shut the system down, wait a minute, and reboot.

The typical $700 HP PC from Circuit City should be shut down every couple of days, especially if kids use it. But I'm sorta hoping the average PC user won't take that advice -- I charge $35 for slow-PC service calls that are cured by rebooting the machine ("Gosh, that's amazing! How'd you do that? And you fixed it so quickly, too! I'm telling my friends! You must be some kind of computer genius!").

If you want a home PC that will hold up to long, hard use, learn to build your own. Or spend $150,00 for a server.

[vcdlover]

Servers are designed to run 24/7 and home PCs are not that's why the servers cost at least 3x more than home PCs. We have servers at our company that cost almost 7 figures which are under a fire proof room. The temp inside the room is set to be at 55F. At least, we have to replace a failure part once a year. All user PCs are recommended to turn off at least on the weekends. The most failure part of our user PCs are motherboards and it seems to be a trend when the motherboards fail. They usually fail when the PCs are turned on. The second most fail part is the cooling fan.

vcdlover

[ronnylov]

pros:
-needed for HTPC recording or other automated tasks
-monitor, drives and most fans can be turned off when not needed using power management

You don't need it running all the time for automated tasks.
In Windows task scheduler you can set an application to start the computer when the program is supposed to run. This works well if you power down in hibernation mode. With my TV recording software (DVBViewer) I can set it to go back to hibernation after the recording is complete. So my PC powers up from hibernation automatically at pre-programmed scheduled times , records the TV-show and goes back to hibernation completely automatically. I use the same method to start antivirus scanning each week. When the PC is in hibernation mode there are no fans running and hard drives are powered down. Well it seems everything is powered down just like a normal shut off, but at least the scheduler must be running somewhere inside the PC, otherwise it would not work.

[oldandinthe way]

Could the heating and cooling of cycling the power on and off not cause solders and board level components to crack and seperate? I know in mechanical applications it is ideal to have a constant temprature...cold starts with quick run-up's cause fatigue and wear, much like a car engine in winter...

No more than a car. The parts are rated for ups and downs -- expected, actually. It assumes a curve of heat-up and cool-down times. It doesn't go from 70-100 in a second, it takes time that is allowed by the parts in use.

A car which is frequently started and stopped, and driven for short distances will suffer far more mechanical failures than one which is driven for prolonged periods at highway speeds.

[oldandinthe way]

With the possible, and I say possible, exception of improved cooling, I know of no design for Intel based servers which is more reliable than most home PCs. And frankly, I doubt that blade servers are cooled as well as some home PCs.

Before the Internet became a major component of information processing in large corporations, almost all of the web servers were recycled, obsolete, desktop computers. Hardware reliability in 24/7 operation was not a major concern.

[lordsmurf]

A car which is frequently started and stopped, and driven for short distances will suffer far more mechanical failures than one which is driven for prolonged periods at highway speeds.

That's not true either. Another BS myth that's often repeated.

[Lucifers_Ghost]

A car which is frequently started and stopped, and driven for short distances will suffer far more mechanical failures than one which is driven for prolonged periods at highway speeds.

That's not true either. Another BS myth that's often repeated.

Says you.

Its a myth that isnt 100% accurate but isnt wrong either.

Like someone said, its 6 of 1 and half a dozen of the other. Cars that are repeatedly started and stopped and drive short distances will have more problems, on average, with their starters and exhaust. Brakes will wear out quicker. The flip side is that cars driven all highway will wear tires and belts quicker.

[edong]

People have different experiences. I'm still using a clone PC bought in 2000 which is being turned on and off at least 3 or more times a day depending on the need to use it. We use this daily because it is located in the family room. My wife will use it for e-mail and internet gaming. I use it for some accounting and video works. We use it like a TV in a sense that, we turn it off after each use.

We have two other PCs upstairs that are being used the same way. Maybe turning on and off very often is bad for anything electronics but my vote is still to turn them off after each use.

[Video Head]

Could the heating and cooling of cycling the power on and off not cause solders and board level components to crack and seperate? I know in mechanical applications it is ideal to have a constant temprature...cold starts with quick run-up's cause fatigue and wear, much like a car engine in winter...

No more than a car. The parts are rated for ups and downs -- expected, actually. It assumes a curve of heat-up and cool-down times. It doesn't go from 70-100 in a second, it takes time that is allowed by the parts in use.

My car is fine in winter, unless it's 10-F (at which point I have to take a hair dryer outside because the stupid doors freeze shut!)

I know it used to be a problem with the constant heating and cooling. Good to hear that the manufacturers have addressed it.

I hear you with the car doors...my car stays inside a heated garage overnight but during the day in -30 weather the doors can get frozen stuck just going to the mall. I use the hipcheck method. People in So. Cal must think we all drive electric cars with the plug sticking through the grill.

[TBoneit]

I do power my machines off when not in use. I pay for electricity.

If I were in the OPs situation where I did not pay for energy and the energy I used was hydro, I would leave my machines on.

If people who want to do what's "green" do so they should be willing to accept as their satisfaction the "virtue" of the deed. They should not expect that it is also better for their equipment.

As for cleanliness, it is next to godliness and I believe in vacuuming out my systems twice a year. Even if they do not run 24/7. I also lubricate my fans at the same time.

Forget Green, It is the cost of the censored electricity. Just because the OP called it Hydro does not mean it is Hydro as such. How do you lubricate your fans? Are you saying twice a year you pull all the fans, clean them, and pull the seal over the bearings and lubricate? Then reinstall. I doubt you'll get Joe Average to go through that based on what I see at work.

Fire is indeed possible, I've seen the results twice. One was due to lightning strike, the network card exploded and caught fire. Second fire originated somewhere in the PC, was found by owners within mnutes as the wallpaper started to go.

It is well known among technicians that the most common failure timing is right after turning the PC on. Surge protection and dust removal are important, but the correlation between hitting the power switch and something going "pop" is clear.

The numbers are not based on a few experiences, but drawing from a large database.

As for the energy savings, a PC at rest equals an average lightbulb or two.

Too True, I've had many a computer brought in where the customer says "I turned it and heard a pop and it's dead now. Too many say they tried turning it on several times. Or they say and I've seen it too, I turned it on and a flame shot out the back (Power supply fan opening.) Sometimes it will be a capacitor on the Motherboard blowing the casing off. Those we do replace.

Some flakey motherboards we look, we see bulging capacitors, case closed repair needed. Depends on teh situation whether we replace or repair. Sometimes you have to repair due to special circumstances. Same reason I keep a couple of used Mobos around that have ISA slots. We have businesses that need them to avoid buying very expensive PCI replacment cards.

Myself I come down on the turn it off. As the way I see it on or off is 6 of one and 1/2 a dozen of the other. So I'll save the money. Plus a computer that is off can not be Zombied.