[WEB4LIB] Re: Powering down PCs

[Richard Wiggins]

Much as I hate to disagree with Walt, I'm going to. :-)

While I agree that this discussion is now 15 years old, and while I agree
that the worries over power cycling a PC are mostly misplaced, I think it's
important to realize that shutting down banks of PCs doesn't answer the
whole problem.

The fact of the matter is that PCs spend a lot of time idle during the work
day, and many work days may not begin and end precisely on a schedule -- and
many banks of PCs are not assigned to just one worker on one shift. So
there's lots of idling during "work" hours.

So... I think the thing to do is to include Energy Star compliance in your
fleet purchase evaluations, and pay particular attention to the monitor's
power consumption when it is in full sleep mode (NEC for instance has
monitors that sleep at 8 watts) and make sure power management is enabled on
all PCs -- and not thwarted by screen savers.

Of course if you've got a large bank of PCs you know will be absolutely idle
over an extended period it may make sense to shut them down. But enabling
Energy Star on every PC could make as big a dent in power consumption.

By the way, a commercial company, Parabon, is going to take a page out of
SETI's book and pay people and institutions to use idle cycle time on their
fleets of PCs. It's really a very neat idea -- just as Gnutella and pals
merge the file system of everyone's PC into a global collective, these folks
want to do the same with spare CPU cycles, and build a global supercomputer
out of millions of PCs.

See http://www.parabon.com/

I'm waiting for some enterprising unit at a university to sign up all their
PCs for the program, only to find that the power plant (which doesn't charge
for electricity) is upset, and the general fund wants the proceeds to go to
it, not to the forward-thinking unit...

/rich

----- Original Message -----
From: <Walt_Crawford at notes.rlg.org>
To: "Multiple recipients of list" <web4lib at webjunction.org>
Sent: Wednesday, August 30, 2000 12:50 PM
Subject: [WEB4LIB] Re: Powering down PCs


>
> Much as I hate to disagree with Dan, I'm going to.
>
> <Soapbox>
> To the best of my knowledge (and I've been following this stuff for 15+
> years now), there is NO credible evidence that powering down PCs (when
> they'll be out of use for periods of more than an hour or two) will
shorten
> their life in any real-world sense.
>
> No PC manufacturer that I know of makes such claims. Nor does any hard
disk
> or display manufacturer, as far as I know.
>
> No body of statistical evidence supports such claims.
>
> PC power supplies are designed to minimize the so-called shock of powering
> up. Hard disks, power switches, video monitors, and other "susceptible"
> components are rated for, at minimum, hundreds of thousands of power
> cycles. Contemporary hard disks aren't threatened by power-down/power-up
> cycles. And so on...
>
> Powering down unused PCs saves power (absolutely). That may not be a
reason
> in areas of unlimited free power generation, but it's a reason everywhere
> else, in terms of both direct and indirect costs. Each idling PC may only
> use 20-40 watts, but that adds up real fast. (A display without
EnergySaver
> settings enabled is an entirely different and much more serious matter:
> you're talking about a *lot* of energy use for no benefit.)
>
> If you're planning to keep your PCs for 25 years or more, there might be
> some bizarre case that burning all that extra power makes them last
longer.
> Otherwise, they should be turned off for extended periods of disuse. (And
> if you are planning to use PCs for 25 years or more, you have bigger
> problems than PC life expectancy.)
>
> Given that (a) the national power grid is already near its limits, (b) in
> some regions [here, for example], we're already facing rolling brownouts
on
> certain days, (c) almost all power generation involves global warming and
> pollution issues, I think that any advocates of leaving PCs (or other
> devices) on permanently need to cite hard evidence. And I don't believe
> such evidence exists.
> </Soapbox>

Richard Wiggins
Consulting, Writing & Training on Internet Topics
www.netfact.com/rww         wiggins at mail.com
517-349-6919 (home office)  517-353-4955 (work)  
______________________________________________
FREE Personalized Email at Mail.com
Sign up at http://www.mail.com/?sr=signup