Every milliwatt of battery power is precious. That's why some shy away from fast-starting, but seemingly battery-draining, "Sleep" or "Suspend" modes on laptops. We measured the actual drain on MacBooks, Windows laptops, and netbooks to determine when sleep makes sense over shutdown.
The Question of "Sleep"
Batteries are filled with mystery, and for good reason—everyone's use of their battery is different, and some batteries are built differently. Some laptop owners still get decent life out of their three-year-old standard battery, while others can barely live through a rapid un-packing of their charging cord once their charge icon turns red. Even the seemingly basic question of whether you should keep your laptop plugged in, or only plugged in for charging, eluded our own Gina when she tried to answer the question for everyone. No two battery abusers, it seems, are the same.
To thicken this mysterious brew of battery knowledge, modern operating systems are pushing the use of "Sleep"—also labeled "Suspend" or "Standby"—over the "Shutdown" we've come to know. It doesn't draw nearly the same kind of power as a fully operational computer, but it does allow you to start your system back up and get back to the things you had open in a few seconds on a relatively new system, rather than wait for everything to get back to stability.
Need a primer on what we mean by sleep, shutdown, and hibernate? Wikipedia's "Sleep mode" entry is helpful, and Microsoft also explains the advantages of sleep.
Modern laptops and netbooks are designed to elegantly fall into sleep and show you their status with blinking or fading indicators. Windows Vista has sleep as its default action on the "Power" button, but changed it back in Windows 7, likely due to (valid) user complaints. And the huge, 7-8-hour batteries being packed into MacBooks are certainly due in some part to the preference of many users to close the lid rather than shut down.
Should you be relying on sleep over shutdown? Will it prematurely lower your battery's total capacity? How can you tell how much juice you'll have left if your laptop sleeps for hours? To answer these questions, a few of your Lifehacker editors put their laptops, and a spare netbook, to the test.
The Tests
These tests are not scientific. I lack for a rig that can monitor voltage and other variables in a sleeping computer. What I can offer is two standard laptops, monitored closely with a timer and battery measurement tools, and some simple graphs of those measurements.
We used two free battery monitoring utilities that offer extensive power details: BatteryCare for Windows, and coconutBattery for Mac. Both are great applications to grab if you're interested in how much your battery holds, how much it used to hold when it came out of the factory, and other data points.
BatteryCare measures charges in milliwatt-hours (mWh), while coconutBattery measures in milliampere-hours (mAh). By dividing the milliwatt-hours by the voltage of the computer they were measured from, though, you get the milliampere hours. Voltage varies in a device—between 11.2 volts on a Sony Vaio and 12.4 volts on a ThinkPad at its peak—so the lowest voltage was used when there were discrepancies; we figured we'd tend toward a worst-case battery use scenario. If our numbers and explanations don't quite meet IEEE standards (they don't), we'll fall back to stating that we took timely measurements of battery strength at each interval, and tried to plot them accurately. Big thanks go to Ray West for his helpful treatise on the nature of batteries.
We measured our portable computers' battery use while sleeping, and with a full shutdown and startup, across a range of thumbnail time periods—one hour, two hours, then four, eight, 12, and, in one test, 24 hours. In all cases, actively open programs were shut down before a sleep or shutdown. Batteries lose charge even when they're not doing anything—just sitting in a shut-off computer—so the strain of shutting down and starting up a system doesn't make it an obvious choice, as you'll see below.
We didn't test hibernation times because, generally, hibernation is a shutdown in which the computer's current state is written to the hard drive, and it draws nearly the same amount of power as a complete shutdown. If hibernation seems to start up again quickly on your system, it, too, can be a smart pick, as can setting your power manager to turn sleep into hibernation after a set number of hours. Your mileage will vary with hibernation, though—I've found it not as reliable, and slow, on my main laptop.
Finally, a note about the MacBook Pro—I was testing it "out of the box," so to speak, without having changed any of its power preferences. Newer Macs can go into "deeper" sleeps, with a mix of hibernate-like disk writing and RAM-stored suspension, and that mix can be modified by utilities like SmartSleep. For the purposes of this test, though, I'm using the stock close-the-lid sleep mode, which restored very quickly after even the eight-hour nap.
The Results
The primary test involved two different computers. One is my trusty Lenovo ThinkPad T61p, running Windows 7. The battery I was testing, however, is an extended life, nine-cell battery I ordered almost exactly one year ago (from LaptopBatteryExpress), that I've only used when there's no power cord to be had, keeping my nearly-dead stock battery in at all other times. The other is the battery inside the brand new 15-inch MacBook Pro.
I put both the ThinkPad and the MacBook through one, two, four, and eight-hour sleep cycles, and measured the ThinkPad after shutting down and starting up after eight hours. The MacBook eight-hour shutdown measurement will be added in later today; it's missing due to a really dumb measurement-keeping error.
Here's how much charge the ThinkPad and MacBook lost over the course of eight hours of sleeping, and how the ThinkPad fared with a shutdown and startup.
Click any of the images below this point for a larger, more clear view
I was surprised at the MacBook's greater use of power than the ThinkPad after the first hour, but also surprised at how both laptops drew on power so consistently. From this chart alone, you might think that shutdown is still the safest choice when it comes to saving on battery power. But let's put those mAh loss figures in perspective.
Here's the same battery drain figures, but measured against a Y axis that uses the maximum charge of the MacBook Pro battery, about 6900 mAh, as perspective:
If you've got to get back to work later in a day, and you've got a decent enough battery, sleeping seems to make some sense.
But those are fairly new and well-kept batteries. How about laptops that have seen some real use? Erica Ho offered to measure up her Sony Vaio and Dell Inspiron laptops after one-hour sleeps and shutdown/startup cycles (her laptops can't really live for four hours on sleep), and the results were, to say the least, surprising:
We triple-checked Erica's figures and organization, but you're looking at the results: shutting down and starting up her old Vaio actually uses more power than letting it quietly sleep. We've certainly seen our share of older hardware that seems to require enough time for tea water to boil before it's ready to go. It's surprising to see that sleep mode, when it works, can make a lot more sense on those systems.
Jason also offered up his Eee PC netbook for a 24-hour sleep test. He didn't have shutdown figures for each marking, but he did learn that his netbook draws power on a very consistent basis, and might be best used sleeping during the day, then charging at night. As he put it, "A whole day of standby is only one-sixth of the battery. Even if I had a bigger and more power-hungry laptop, though, I'd still use standby and just be extra-sure to plug it in at the end of the day."
As stated at the beginning, each computer user's habits are different, and each computer user has different scenarios to work inside. Still, from testing a few modern and not-so-modern laptops, we've come to see sleep as a pretty smart option, if you know you won't be leaving your system alone for more than a day.
Have you done your own formal or informal measurements of battery drain with sleep versus shutdown? Found a completely different experience with your own portable systems? Tell us your take in the comments.