Home Forums FineTunedMac Community Lounge Should I turn my Mac off or let it sleep?

Continuing a discussion that began here...

dkmarsh has posted that

I think people would be better able to make use of that information if they also knew how much power those Macs draw when they're sleeping.

Also, my Early 2009 White MacBook draws 60W, but how much power does it draw when it's sleeping, how much of that would the power-block be drawing under any circumstances, even if my deuced Mac(hina) was shut down, and does battery degradation enter into the equation?

Thanks.

Edit: Wouldn't it make sense to detach the linked series of posts and move it here?

7.5 amps!!!
Shocking.

I'm revolted. My ohmy. There seems to be an infinite capacity for the puns that are making the circuit.

Now we know why your neighbors have isolated you and won't let you off the property.

Watt did you say? For the record, my wife claims that she married me in order to save the world from having me around. I don't know why I married her because I vowed never to marry the type of woman who would have me for a husband.

Those figures represent "maximum continuous power." I'm not sure how that compares with, say, typical usage; during maximum power consumption, we'd expect the processor(s) to be fully loaded and the fans ramped up, which isn't typically the case for most of us, I think.

Note that the discussion you linked to isn't specifically about whether to turn your Mac off or let it sleep; it's about whether or not to turn your Mac off, period. Energy usage may be one factor taken into consideration when deciding whether to shut down regularly or not, but I have no idea how to derive power consumption during sleep from "maximum continuous power," so I can't help folks weigh sleep against nonsleep. My referencing the mini's power consumption was simply to illustrate that its energy usage is so comparatively slight that I don't really have to take it into account.

> Those figures represent "maximum continuous power." I'm not sure how that compares with, say, typical usage; during maximum power consumption, we'd expect the processor(s) to be fully loaded and the fans ramped up, which isn't typically the case for most of us, I think.

That gives your original post both context and perspective.

> My referencing the mini's power consumption was simply to illustrate that its energy usage is so comparatively slight that I don't really have to take it into account.

As I said, perspective... A continuous 85 watt power draw comes to 2kWh/day, about 50% of my average daily usage (other than in A/C season).

> Note that the discussion you linked to isn't specifically about whether to turn your Mac off or let it sleep; it's about whether or not to turn your Mac off, period.

Well... Since leaving your Mac on and unused frequently leads to sleep, and despite the fact that power consumption was hardly central to the original discussion, power consumption should enter into any "on/off" discussion, particularly considering the MDD's (among others' ?) relatively humongous draw and the wave of "green" engulfing the US.

Wow! My hat's off to you, sir. I figure it takes me about 3.5 kWh to bake bread, which I do about twice a week, so that's a quarter of your energy consumption right there.

It appears that, as expected, the maximum continuous power figures are rarely if ever attained in actual use; various anecdotal evidence suggests cutting those figures by one half to three quarters will result in a more realistic estimate. In the case of the mini, Mac Mini: Power Supply yields some interesting observations. The figure of 25 watts recorded by Boyd Waters for idle is twice the value found by Best energy saving Apple Desktop Computers; differing definitions of idle may be the variable responsible. (I'll note that when I'm not at home, and overnight, my Mac is set for display sleep, so any power required for the graphics card to draw to the screen is negated.)

In any event, it's clear that for all models, the difference between off and sleeping is a very slight fraction of the difference between either and on, even at an idle, which would suggest that power consumption is not much of a factor in deciding which of those two options one chooses. I guess that was really my original point: if one wants to keep one's Mac awake 24/7, then power consumption becomes a (more) relevant concern.

FWIW, my Samsung LCD display consumes 34 watts when on, and the same nominal 1 watt whether asleep or off. So my mini, during the 85% or so of the day that I'm not using it, uses, say 15 watts, and the display uses 1 watt:

.85 x 24 x (15 + 1) = .326

meaning I'm "wasting" 120 kWh per year by keeping my mini awake 24/7, or about the amount of power consumed by a medium-sized room air conditioner running six hours a day for three weeks. (And the actual figure is probably lower, because booting up—which I'd be doing probably half a dozen times a day if I were shutting down when done with each "session"—appears to be an energy-intensive activity which I avoid altogether now.)

Nice analysis which, I think, does a pretty good job of fleshing out the power usage aspect of this discussion.

Thanks.

(The only question left unanswered, which is only marginally relevant to this discussion, is how much power a power-block draws just being plugged into the wall? [If Google knows the answer, either it isn't telling or I asked the wrong question.])

As a completely side note, the whopping 7.5 amps the MDD G4 systems draw (and the even more prodigious power requirements of the G5 systems) are one of the main reasons we're using Intel Macs today. IBM wasn't terribly interested, and Motorola didn't seem terribly able, to get the power footprint of the PowerPC chips down to something reasonable, and that, more than performance, is what drove Apple to Intel, from my understanding.

The PowerPC architecture was originally intended for minicomputer, workstation, and supercomputer applications, where power efficiency isn't a top requirement. The IBM BlueGene/L uses 133,120 PowerPC 440 processors (a PowerPC 440 is basically a G4 with special I/O and optimization for multiprocessor mainframes); each BlueGene/L processor module is two PowerPC 440s, and there are 65,536 main processor modules and 1,024 processor modules dedicated to getting information into and out of the beast. Power consumption is 25 kW per rack times 64 racks for a whopping 1.6 megawatts of power...a figure that's so steep that the later IBM BlueGene/P and RoadRunner supercomputers used processors optimized for lower power consumption.

A part of me wonders, if Apple had stuck with PowerPC processors, if we'd be seeing some of the work that was done to make the BlueGene/P processors less power-hungry trickle down to the hypothetical G6 and/or G7 processors. Moot point now, I suppose.