Slashdot Mirror
Why Does Windows Have Terrible Battery Life?
An anonymous reader writes "Jeff Atwood at Coding Horror is trying to figure out why the battery life for devices running Windows is so much worse than similar (or identical) devices running other operating systems. For example, the Surface Pro 2 made great strides over the original Surface Pro, increasing web-browsing battery life by 42%, but it still lags far behind Android and iOS tablets. The deficit doesn't get any better when Windows is run on Apple hardware. Atwood says, 'Microsoft positions Windows 8 as an operating system that's great for tablets, which are designed for casual web browsing and light app use – but how can that possibly be true when Windows idle power management is so much worse than the competition's desktop operating system in OS X – much less their tablet and phone operating system, iOS?' Anand Lal Shimpi is perplexed, too. Atwood is now reaching out to the community for answers: 'None of the PC vendors he spoke to could justify it, or produce a Windows box that managed similar battery life to OS X. And that battery life gap is worse today – even when using Microsoft's own hardware, designed in Microsoft's labs, running Microsoft's latest operating system released this week. Microsoft can no longer hand wave this vast difference away based on vague references to "poorly optimized third party drivers." ... I just wish somebody could explain to me and Anand why Windows is so awful at managing idle power.'"
...because it's old and bloated!
Watch the WWDC sessions on power management in iOS and Mac OS X. You'll get an idea of how much work Apple put into this over the last decade or so.
-jcr
The only title of honor that a tyrant can grant is "Enemy of the State."
Sheer love of evil! Seriously, though - all the massive background processes. Probably a decades-old stack of services and whatnot they don't have the corporate continuity to be able to change at this point.
Ah ha: "I just wish somebody could explain to me and Anand why Windows is so awful at managing idle power."
You make the mistake of thinking that just because the device isn't doing something at the user's direction, that it is idle. How do you think the NSA is getting all of their number crunching done while they shake the bugs out of their Utah data center?
I am not aware of a phone yet that can run more than one operating system. Comparing a windows phone to an iPhone is a pretty useless thing to do as their hardware is quite a bit different.
Damn_registrars has no butt-hole. Damn_registrars has no use for a butt-hole.
Much like Jay Leno, they've built up a LOT of legacy bloat over the decades.
The cow says "Moo." The dog says "Woof." The Timothy says "Thanks, valued customer. We appreciate your input."
Not in my experience. Over the past twenty years I've run Linux on a large number of designed-for-Windows laptops; I've never seen worse battery performance under Linux than under Windows, and on some machine (including my current Asus Zenbook) considerably better.
I'm old enough to remember when discussions on Slashdot were well informed.
Someone should introduce them to my laptop running linux!
Don't complain about syntax, grammar, or spelling. There is no.hell like input on android.
The fact that you can do two things at once doesn't mean that you have to do lost of pointless work all the time.
A Pirate and a Puritan look the same on a balance sheet.
A few questions that would be interesting to know the answers to:
- Is the power consumption deficiency the same across all hardware or does it close the gap on certain pieces of hardware?
- Is the consumption deficiency gap the same on tablets vs laptops vs PCs?
- How much can Windows 8 be tweaked to save battery life (IE: disabled unneeded services)?
- Does it manage power of certain pieces of hardware better than others (SSD vs HDD, AMD vs Intel)?
- Do drivers make a difference in power consumption?
- How many hamsters have heart attacks every time Windows 8 is benchmarked?
> Why do Windows 8 tablets running on an Intel i5 x86 CPU
Actually, the newer Intel CPUs are supposed to suck a lot less when it comes to power consumption. Atoms just suck in general. The fact that they don't draw any power is hardly anything to write home about.
A Pirate and a Puritan look the same on a balance sheet.
Funny you should mention Symbian - my S^3 phones had the best battery life of any of my smartphones, regardless of platform, without having larger batteries than their iOS/Android/Windows Phone counterparts.
Comparing Android or iOS on ARM to Windows (or OSX or Linux) on a full i386 platform is simply meaningless. Why do it?
The (linked) Aandtech article on battery life pretty much answers its own question.
Surface pro and surface pro 2 completely destroy everything else in the benchmark ratings. It means haswell doesn't manage lower power scenarios nearly as well as ARM, but Intel never has.
For a comparison to iOS they'd need to well, actually have on on their chart. I can certainly see the argument that Windows is worse at power management than other OS's on the same hardware - but without hard numbers in a chart that's a tough case to make, since you're comparing different review sites to each other. Comparing different hardware is missing out on a lot - for most computing needs they're benchmarking Haswell is massive overkill - which might just be it, it literally cannot slow itself down enough (with either MS or intel drivers being the culprit) to save even more power.
Or windows is doing background stuff that other OS's aren't. Whether those provide any value to justify reduced battery life or not is debatable, but the answer seems to be 'probably not'.
It still isn't 'microsofts hardware', it's hardware from some 3rd party vendor they soldered together in a case and put their own sticker on it. Yes, it's up to MS to try and ride the cases of Intel and whomever is supplying their displays and SSD's to find ways to save power, but it's ultimately up to the 3rd party guys (who also sell parts to the rest of us) to actually make the drivers for their hardware.
It looks like some people missed the part near the end of the post where they explain how Windows power usage is abysmal even compared to the desktop version of OS X.
It looks like some people missed the part where no tablet runs OS X.
I know that my virus scanning service seems to be running at 2-5% most of the time. And, my process list looks a mile long.
I think we expect our windows devices to be real computers and load them up with full applications. Then, we expect them to sip juice like Android. Can't comment on the OSX. My netbook on linux is 5 years old and doesn't have much of a battery left.
Where all think alike, no one thinks very much.
battery life was absolutely terrible for me w/ linux. measured in minutes not hours. windows was at least 150% better on the same system (which still wasn't great).
Chewing up all your CPU.
(Yes, this post is a joke. It's an (in)famous old article from everyone's favorite tech writer -- who was, in fact, being serious.)
Dear Slashdot: next time you want to mess with the site, add a rich-text editor for comments.
The author is a massive troll for comparing Surface Pro hardware (which runs a full blown i5 processor) with iOS and Android hardware (which is typically far lower power both in terms of wattage and processing).
He's also comparing it against a MacBook, which can have exactly the same i5 processor. See the part in TFS about how running Windows on Apple Hardware doesn't actually change the deficit?
Pretty sure Windows generally gets (sometimes substantially) better battery life than Linux.
Depends on what you're doing. My laptop gets better life on Linux than it ever did in Windows, but all I do with it is surf the web. It doesn't require a lot of processing power, and Windows wastes a lot of clock cycles running stuff it doesn't need to accomplish the task.
Only on Apple hardware, which requires Apple drivers for power management, and surprise surprise, Apple sucks at Windows drivers (and always has). In one particular, the Windows power management drivers for my friend's MBP don't suppose variable fan speed control. It always runs full speed. No shit, that's going to waste battery life... On the flip side of the coin, though, Hackintoshes get worse battery life than Windows on the same hardware. This entire "article" is stupid; anybody who isn't blinded by fanboyism and has used the systems in question could tell you that.
Surface Pro [2] has worse battery life than an iPad or Android tablet for a simple and bloody obvious reason: Core i5 CPU. Not some power-sipping little ARM chip with passive cooling, but full laptop-grade 64-bit processor. Even completely leaving aside the obvious (to anybody who is not an idiot, which apparently excludes the submitter) differences between a desktop OS (Win8.x) and a mobile one (Android or iOS), there are very obvious reasons for the battery life difference.
There's no place I could be, since I've found Serenity...
So much so, that when I fire up XP it goes into TURBOFAN MODE and CPU temps still climb into nutsack-roasting level. 90 to 100 C for the CPU temps (Core2 Duo) have occurred without too much heavy lifting. So forget about the battery life, there is no use without the power cord. It's more an issue to be concerned with the physical limits of the rest of the hardware, like when does it melt?
Read down a little further, he compares an MBA and a Surface Pro 2 running anad's wifi web browsing benchmark. The hardware is very similar, but the MBA lasts about twice as long.
Comparing operating systems running different hardware is a meaningless endeavor.
Comparing on the same hardware is better until you take the logical leap of drawing general conclusions from it.
When you use the conclusions above to draw additional conclusions about what you think would happen your ability to predict or be taken seriously takes a hit.
My 5 year old lenovo draws ~7 watts on battery with the 14" display on and 7200 RPM platter spinning. I am able to observe consumption difference from battery manager in detail when I turn hardware on and off.. run applications..etc.
The answer is likely knowable if only there was willingness to spend more time (thinking), measuring and working the problem and less time (talking) drawing conclusions.
Errm, what article are you reading? Because the one I see shows the Haswell-powered 13" MBA getting ~14 hours of battery life to the Surface Pro 2's ~7 hours of battery life. Sure, the 13" MBA has a bigger battery, but the 11" MBA has a smaller battery and still gets ~11 hours.
Your arguments about the Surface Pro 2 not really being microsoft hardware are not really meaningful, you could say the same about Apple's notebooks. They don't make the CPU, or the GPU, or the SSD controller, or the screen, or the display controller, etc.
Comparing a Surface Pro to an iPad is about as useful as comparing a bus to a small car. Of course the small car uses less fuel, but I'd like to see a bus full of people crammed into said car.
The iPad and the tablets that it inspired are the new netbooks: barely useful for anything beyond simple tasks.
The Surface Pro and similar tablets are ultabooks stuffed into tablets - this has advantages and disadvantages.
As for OS X, that is indeed somewhat misterious, but it probably boils down to:
- Driver optimizations: having a very limited set of hardware that needs to be supported makes it much easier to optimize drivers (and if needed the OS itself).
- Bloatware: my Ativ Smart PC Pro came with at least three Samsung applications that constantly run in the background and (way too often) interact with the user. Control panel thingies for this and that driver don't help, either. Some of those probably misbehave and screw up the scheduler enough to measurably reduce idle time. These are not present on OS X.
- UI: I'm not sure just how much hardware acceleration OS X uses, but Windows Vista/7 with Aero and Windows 8 at all times have hardware accelerated graphics for their UIs - eye candy in exchange for power consumption.
- Unusually low-power hardware: I can imagine Apple applying pressure for individual components' power consumption to be lowered - the screen comes to mind as a likely culprit.
(1) there is so much cruft under the surface in Windows (fake DOS calls, umpteen levels of virtualism, etc) that the machine expends a ton of cycles doing what is NOP in newer systems not supporting 1980 calls.
(2) optimization isn't pretty and doesn't sell, so Microsoft is not cleaning house.
if this is supposed to be a new economy, how come they still want my old fashioned money?
Defragging a potentially huge disk, in the background, on-the-fly, so the disk never slows down.
File search index, in the background, on-the-fly, so you can search faster. You can turn this off.
Full window dragging, and many other graphics enhancements. You can turn these off.
Is the printer still there? Let's check again.
Port polling, did you know that a USB port might gett polled 50'000 times per second? You can turn this down. A lot.
Scheduled tasks. Oh so many scheduled tasks. You probably have over 1'000 defined.
Is the internet still connected? Let's check again.
An actual software Firewall. You can turn it off, or make it much simpler.
Multi-user, multi-profile. Everything gets doubled.
Is the printer still there? Let's check again.
Is the internet still connected? Let's check again.
Event logging. Windows knows what it's doing, because it takes the time to write it down.
The windows registry. It's probably the single most reliable aspect of any operating system. It's incredibly fast, always-on, used tens of thousands of times in a single moment by a any application -- my graphics suite writes 12'000 registry entries when I close the application. And you never need to worry about it getting corrupted.
No fewer than eight different scripting languages available at any moment.
Twenty versions of a single DLL loaded concurrently, for cross-decade application compatibility.
It's not just an operating system. It's a generic operating system that can run anything from decades ago. My 1985 application still runs on my vista machine, which is still running smoothly 7 years after I built it, and now it's running software 7 years newer than it is. iOS doesn't do that. Neither does OS X. Neither does Android.
But there's always been a version of windows with better battery life. It used to be called XP embedded. And it was exactly what you expected it to be -- you got to just start turning off huge parts of windows. You're welcome to do it. No, you don't want to. You don't want things to be slower, and you don't want to lose all of those great features. And many are tied together.
And that's why you chose a windows machine in the first place. Not because it does the bare minimum, and hence saves battery life, but because it does everything it's always done at a reasonable battery life.
But hey. If you want to complain about power vs features, I want you to look at my tvision's on-screen menu system. Now it's a smart tv, with a menu of icons to all sorts of dumb shit. And yet, just scrolling through those pages of icons is slower than my speak'n'spell. My tvision is plugged into the wall, with as much power as it wants. The led light bulb consumes more power than the computer running the on-screen menu. Why? I have no idea. But it also doesn't have a pre-amp, so I can't plug in any headphones or larger speakers without an optical cable and a home theatre amp/receiver. Thanks for that.
Not in my experience. Over the past twenty years I've run Linux on a large number of designed-for-Windows laptops; I've never seen worse battery performance under Linux than under Windows, and on some machine (including my current Asus Zenbook) considerably better.
Linux didn't run on many laptops in '93... and for a long time even this milennium, Linux on laptops was painful because of partially functioning hardware (docking and screen mirroring/dual screens), poorly working suspend/resume and poor battery life. While Linux was great for workstations - and by far the best choice for servers - let's not paint too pretty a picture. Working well on laptops is a far more recent addition.
Moreover, the Sony Vaio Pro has a higher-resolution screen than the MBA, which puts the Vaio at a disadvantage (because it drains the battery a little faster). So with highly-optimized Windows drivers, the battery life looks the same or even better for Windows.
The comparison to ARM is just stupid. Obviously battery life is better on ARM, at the cost of much lower performance. That's true for Windows and OS X both.
This happens because you don't have the correct Intel thermal profile driver installed. There won't be any unknown devices in Device Manager, but you still need to install it or you will see the behaviour you are getting in Windows.
Intel allowed manufacturers using C2D processors to customize the thermal profile of the CPU based on what their hardware was capable of. That's how very thin and light laptops were able to use high end CPUs without overheating, but limiting the amount they can ramp up CPU speeds and voltages, particularly when both cores are loaded or the GPU is also active.
Apple must provide the correct drivers, tailored to their laptops. Presumably you can download them from somewhere.
const int one = 65536; (Silvermoon, Texture.cs)
SJW, n: "Someone I don't like, and by the way I'm a fuckwit" - AC
Pretty sure Windows generally gets (sometimes substantially) better battery life than Linux.
Rubbish. The netbook I'm typing this on right now dual-boots Ubuntu (full Unity bloat version) and Windows 7 Starter. It gets (or used to get a year or so ago, when the battery was a little newer) 6.5 hours under Ubuntu, more like 4.5 under Windows.
Windows just isn't built to be light. It tries to do A LOT in the background to "improver your experience". Some of which might even work. But a lot of it will turn out to have been wasted effort ("Wow, you've indexed all the files in my Program Files folder, well done! If only I had any reason to access a single one of those files today..."). And in the meantime, you've managed to consume a full third of my battery...
There are lots of posts here claiming that they aren't using like to like comparisons. The point of the post isn't that an iPhone is getting better battery life than a MS Exchange Server, the point of the article is that in almost every scenario you can match up, Android/iOS/OS X comes out clearly ahead. And this is the case regardless of what hardware or type of hardware you're comparing. Put Windows on a MacBook and it's going to get lower battery life... a-ha, it's a driver issue, you say, ok, but spec out a similar PC notebook and it will have lower battery life than the MacBook.
In other words, Microsoft doesn't have a battery life on the Surface RT or any other product problem, Microsoft has a battery life problem. Why is that?
If a multi-hour Windows battery becomes a multi-minute battery in Linux, you have a serious problem that needs to be fixed. That is not normal system behaviour.
Install problems are rare these days with Linux, but they're not unheard of. I'd guess that you've got some sort of hardware driver issue if the battery is discharging as crazily as that.
My experience of battery life with Linux (mostly using full-fat, full-bloat Ubuntu) is that it has always as good or better than Windows, except on one laptop I once installed on which I had serious driver issues with, which first refused to charge the battery at all, and then (after I'd fixed that) was discharging within about an hour. Once I'd fixed it, though, it was back to normal. And one problem-filled laptop build out of many dozens isn't a bad record really.
Apple wrote ios for devices with batteries in mind from day one.
It is slowing getting close to OSX, but they have to be careful.
Microsoft chose the "one O/S everywhere" serving needs of desktop
and devices at the same time. This was a naive approach.
For a fair comparison, you should make sure to use power management software with Linux too. Jupiter used to be the top favourite, but I believe that's fallen out of use now. I hear TLP is good, but never used it. Both tools should do what the manufacturer's Windows tool was doing- mostly ramping down the processor when under light use.
My Nexus 4 phone barely lasts my work day. My iPad needs to be charged every day. I can get a few days out of my Nexus 10 unless I even touch a game. About the only device that lasts the week is my iPod Touch, but then I use it mostly as my alarm clock.
I think this is a pretty universal problem. Batteries have not kept up to the demand of CPU performance required by our devices, period.
Of course with relevance to article. when the author realizes that Surface Pro is a laptop (i.e. PC ) and iPad is a device built from phone hardware maybe he might realize how stupid the question was.
It would be more relevant to compare Surface Pro to MacBooks and ask how Macbooks can last the day while Surface Pro won't last more then a few hours.
I haven't thought of anything clever to put here, but then again most of you haven't either.
I would have been interested to see what the gap looks like on a Hackintosh where, presumably, hardware optimizations would be in Windows' favor. I suspect the gap would remain, since the battery optimizations don't depend much on device drivers.
You are going at this the wrong way. First off you cant compare a full PC operating system to a mobile device, so lets throw that out of the window. The question is why is OS X so good at power management? Windows is actually not bad at power management, most people here on Slashdot (even die hard Linux users) will admit that Linux is by far the worst out of the 3 on a PC platform. As someone else above me stated earlier, Apple have put a lot of R&D into researching battery run time performance. Their inclusion of two inbuilt graphic cards (one high performance and one for general usage) being a big step they took into improving battery runtime. This is why even Windows on a Macbook will not perform as well as OS X because it just havent got support to switch between two graphic cards based on what the system actually needs at that time. There are obviously a lot more tweaks Apple have done as well, but there are too many to list here...
I used to be fairly familiar with this problem, and the answer is that there isn't one answer. There are lots, which is why it's such a hard problem for MSFT to tackle.
Some of it is hardware choices - Windows is a pretty heavyweight OS, so the tendency is to use Intel instead of ARM, higher CPU clock speeds and larger (S|D)RAMs to make it feel at home. Plus of course the apps are a bloatfest. The result is higher power consumption.
Some of it is Windows thinking it owns the idle time to do whatever it wants (defrag, virus, search indexing, etc.). On other mobile OSs, the idle time is owned by power management first and foremost.
Some of it is crappy drivers that don't bother to take advantage of power management features present in the hardware.
Some of it is Windows not having APIs to make apps power aware from the beginning. As a result lots of things you'd like to do break old apps, and backwards compatibility is viewed as key.
Some of it is irrationality in the Windows hardware logo testing procedures that cause HW vendors to do bass ackwards stuff to make Microsoft happy.
Some of it is that Microsoft didn't push power to the WinTel HW vendors very hard in the past, and it's hard to change directions on a dime.
Some of it is that Windows applications don't really have an API for sporadic updates out of idle timed by the OS (mail fetch etc) and instead have a tendency to set timers and go do things on their own schedule. Since the schedules aren't coordinated, the device isn't idle much.
Add all those problems up, and you have a structural gap that's basically impossible to fix by changing any one thing.
"The windows registry. It's probably the single most reliable aspect of any operating system."
FUCK YOU.
Sorry, reflex action from a decade and a half of dealing with the "most reliable aspect of any operating system" and the thing about windows that really drove me to OSX.
"There is more worth loving than we have strength to love." - Brian Jay Stanley
From the comments on the main article, I read this link [1].
The kicker? This was from 2009, referencing 10.5.7, a four-year old OSX vs. Windows 7. I'd be interested to see if a recent netbook hackintosh with Mavericks vs. WIndows 8.1 would show... likely an even wider divergence given the findings in this /. post.
[1] http://www.mobilemag.com/2009/05/14/hackintosh-netbooks-experience-33-battery-life-boost/
Make sure everyone's vote counts: Verified Voting
Ah, so Microsoft is running the unoptimized Apple drivers on the Surface tablets..
That explains everything!