Windows 8 Consumer Preview: A Quick Look at Battery Life (Updated)
by Jarred Walton on March 1, 2012 2:15 AM ESTWith each new operating system in recent history, Microsoft has promised better battery life. We tested this back in the early days of Windows 7 and found that while Vista was generally a step back relative to XP, Windows 7 fixed much of what was wrong and even managed to beat XP in several tests. With the Windows 8 Consumer Preview now available, we thought we’d run a quick test on a laptop to see if things have changed much. Let’s just get this out of the way, shall we?
Okay, that’s a pretty poor showing, but what’s really going on here? We’re using the same Sandy Bridge laptop that we tested back when Sandy Bridge first launched, a Compal manufactured unit with an i7-2820QM processor, 4GB RAM, and an Intel 160GB G2 SSD. So the hardware hasn’t changed, but battery life is much worse right now—and that last part is important: right now. We did a few tests of battery life with the Windows 7 preview and it didn’t look that great, but drivers and optimizations weren’t finalized, and Windows 8 CP is definitely in that same category. But there’s more to the story than just Windows 8.
As part of the Windows 8 CP experience, we’re also given the privilege of running the Internet Explorer 10 Consumer Preview (IE10CP). As we’ve noted in the past, the choice of browser can certainly have an impact on battery life, and that likely goes double when we’re looking at beta software for the OS and browser. I also performed an in-place upgrade from Windows 7, so it’s possible that could negatively impact battery life as well.
I've only had a chance to run the battery drain test once so far, so consider the above results very preliminary. I'm going to test it again, as well as go back and retest with Windows 7 (and IE9) to see if there are any other changes. The battery may not be performing as well as it did last year (though it seems to be fine based on HWmonitor reporting 2% wear level), but we'll hold off on any final verdicts for now. While I continue to look into battery life over the next few days to see if perhaps I missed something, it doesn’t look like Win8 CP with IE10 CP is going to do anyone any favors in terms of accessing the Internet while unplugged.
Update: In case you need further explanation, the results above are simply a first test of battery life using IE10 CP. I reran the test a second time and it improved slightly (263 vs. 250 minutes), but I'm still testing. There is also a newer Intel driver that I've now installed, which may help quite a bit for IE10. The above results do not say anything about idle battery life, battery life using Metro apps, battery life during video playback, etc. I am working on testing those items as well, and I have a second laptop that I'll be using to provide additional results. All we can say right now is that after first installing Win8 CP via an upgrade to Win7 and when using IE10 with Flash enabled, battery life looks poor—like, Safari browser on Windows levels of poor. That can and very likely will change before the final release, and the fixes will likely come in the way of driver updates as well as improvements to the browser.
Update #2: I've added results from a second laptop where I have done a clean install of Windows 8 CP. This time the laptop is the ASUS K53E with an i5-2520M processor. I did retest with Windows 7 running Internet Explorer 9 first, and I also swapped out the hard drive for a 64GB Kingston SSDNow V100. Results are much better than with the first laptop, but there are a few remaining elements I need to test. Again, consider all these results preliminary, but at least it does appear that Windows 8 with IE10 battery life may not be quite as bad as my initial results indicate. (Note also that using IE9 in place of IE8 actually improved battery life with the K53E and the V100 SSD—Win7 with IE8 scored 333 minutes. I'm not sure if that will always be the case, however, as I seem to recall seeing IE8 get better battery life on at least one laptop I tested.)
Things still waiting to be tested: first, I haven't finished retesting the i7-2820QM with the latest Intel HD Graphics driver [Update #3: the new driver did not change the result], second I need to retest it after doing a clean install of Windows 8 rather than an upgrade, and third I need to retest with Windows 7. [Update #4: I did a clean install of Windows 7 on the i7-2820QM and reran the Internet test twice. The first result was 393 while the second was 425, so other than variance between runs (possibly the fresh install somehow played a factor on the first run), it doesn't look like battery quality has deteriorated. The graph has been updated with the latest numbers.] I'll hold off on reporting idle/video/alternate browser battery life for a future article.
Final Update: After running numerous other tests in Windows 7 just to verify the numbers we had there (and in the process of working to put together graphs for a larger article), I started with a clean install of Windows 7, finished the benchmarks, and then did an upgrade to Windows 8. This time, the battery life is much better--358 minutes compared to 263 minutes from the original upgrade. It's likely that the original upgrade had a lot more stuff left over that somehow impacted battery life. Now, the numbers are much more in alignment with the results of the K53E laptop. We'll have full details on several battery life test scenarios in a future article, but there's still a drop in battery life right now of around 15-17%, mostly likely because of differences between IE9 and IE10CP.
And just as an aside, I know that Windows 8 isn’t final by any stretch of the imagination, but while the Metro UI (and UI changes in general) seems like it would work great on a tablet, I’m ready to go on record as saying I think it sucks for traditional desktop and laptop users. Without a touch interface, Metro feels weird at best and downright awful at worst. What’s more, getting a touchscreen for a desktop or laptop isn’t actually something I’m clamoring for. Does using a 24” or 30” touchscreen on my desktop sound enjoyable? Not at all, and a 15” laptop touchscreen wouldn’t be much better. Maybe it would help me build up a bit of arm strength, but that’s about the only upside. Add on fingerprints—a personal pet peeve that smartphones and tablets still suffer from—and I’m more than happy to stick with the “boring” old Start Menu. That’s just my initial impression of course; anyone else have similar—or different—feelings after playing around with the Win8 CP?
Facebook
Twitter
RSS
113 Comments
View All Comments
quiksilvr - Thursday, March 1, 2012 - link
Seriously, just run Firefox or Chrome and see how they stack up.fishbits - Thursday, March 1, 2012 - link
"I also performed an in-place upgrade from Windows 7"You what?! Really? For testing purposes to be reported to the public? What?"
"I've only had a chance to run the battery drain test once so far, so consider the above results very preliminary."
But publishable, right? And without bothering to remove obvious variables like battery aging or using the same browser?
"...while the UI seems like it would work great on a tablet I’m ready to go on record as saying it sucks for traditional desktop and laptop users"
The UI? Or did you mean the Metro portion of the UI? Because that's a pretty important distinction, I'd think. And that has what to do with a battery test?
This is sloppy and weak, a trend that's been going on recently here. Hope Anand's still on top of things, because what's been going on is a far departure from his methodology and presentation style.
jdmackes - Thursday, March 1, 2012 - link
I'm glad someone else pointed those comments out. If you've only tested the battery once, how can you publish an article like this? You're accusing microsoft and windows 8 of having poor battery life without even doing proper tests.JarredWalton - Thursday, March 1, 2012 - link
One word: pipeline. It's not a full article, and if you'd notice it does say "a quick look". I'm still testing and will follow up with more numbers once I can charge/discharge the battery about six times under different workloads. But thanks for pointing out the obvious concerns that I specifically spelled out in the text and then saying that we're going downhill.fishbits - Thursday, March 1, 2012 - link
Knowing that your methodology was so shabby, by your own admission, you shouldn't have published this until you could do something proper. Not "definitive" or "authoritative" or "exhaustive..." simply proper.If the results were within 10% of each other one could say that's reasonable variance, but that follow-up is to come. When the discrepancy is that huge, most objective testers would prefer getting it right to.... what is your goal? Why would you publish something that *you* say was poorly observed?
And yes, compared to the Anandtech that I've read for many years, rushing to publish results the author knows are flawed IS going downhill. Anand is relentless in trying to weed out his own biases, assumptions and bad observations in his own writing. He isn't perfect, but he sure tries, and that level of effort is appreciated.
JarredWalton - Thursday, March 1, 2012 - link
It's not poorly observed, it's a poor result. And I've rerun the test a second time (so the battery is likely holding a better charge after one full cycle) and the results were 4.8% higher than the first run.I've got a second laptop running the exact same test, this time with a clean install. I've made sure it has the latest Intel HD Graphics drivers (those were missing on the system in this test, so I've updated and am currently running the test yet again). The current estimate is 313 minutes with Win8 + IE10, compared to 360 minutes with the original Win7 configuration. But that's only an estimate based on the "battery life remaining" and "estimated time", so I wouldn't take those results as proof of an improvement until the battery is actually fully drained. It's closer to the Win7 result for sure, but it's also still a 15% drop.
So questions that still remain and need to be answered:
1) Is the poor result because I did an in-place upgrade of an existing Win7 install?
2) Did the outdated Intel HD driver cause a severe drop?
3) Will we see better battery life results with a different browser instead of IE10CP?
4) What about battery life in other scenarios (e.g. idle and video playback)?
5) Does dual-core vs. quad-core CPU use show any relative differences between Win7 and Win8?
There are probably a few other points that we could raise, but those are the main items right now.
toytanks - Thursday, March 1, 2012 - link
I ran two tests with a laptop I bought around Black Friday (Samsung RC501, I think) pitting the DP against Windows 7 SP1. For the first I ran a program called Battery Eater and the second I looped Apocalypse Now on Netflix until the battery died and in both cases the DP lasted longer. Unless they did something godawful in the CP these results are hard to believe.Braumin - Thursday, March 1, 2012 - link
I am so sick of these "power users" who try Metro for 10-20 minutes, and then say they hate it.You know what? I used to hate brussel sprouts as a kid, but I tried them enough that now I love them.
Huge changes like this are going to take some time to adapt to. Give it some time, and maybe you will love it and hate going back.
Every time there is a new version of Windows, it always takes some adjustment before you can intuitvely do what you want. With a change this big, it is going to take slightly longer than 20 minutes.
Do you really think average joe consumer will complain as much as you? They are going to log in to their desktop, and see their live tile for mail, and it will show them they have 10 unread emails. Then, they will see that they have 5 facebook updates on their facebook app. I would argue that average consumers are going to transition to this way easier than so-called power users, who are so stuck in their old paradigms that they don't want to learn anything new.
As for the battery life (what this article was supposed to be about), it is obvious Jarred did not do his homework here. I would suggest he retracts this, and runs a proper set of tests.
mongster122 - Thursday, March 1, 2012 - link
Have you personally tried Metro with keyboard and mouse?My comment with Win 8 is it seems like we're back to Win 3.1 and Win 95 where you have Windows running on top of DOS. Microsoft is touting one single OS for all devices but it doesn't feel like it. For touch devices and less resource-intensive apps you have Metro, for the more resource-intensive and "business" apps you go back to classic Windows. That's the same for IE 10. You get Metro IE and then you have classic IE. Why not just one that can "sense" if it's running in a touch or a desktop environment and launch the appropriate version?
Metro is good but I wish it would run on the classic desktop. I don't want to be having to switch to Metro and back every so often. The Start button doesn't have to completely go. Just get rid of the icons in the classic desktop and replace Metro with it. That way we can get the best of both worlds. I think the only reason MS can't do that is Metro and classic Windows are still two separate entities and not one seamless OS. That way they can disable classic Windows when running on mobile devices that need Metro only. Another way of saying it is Windows 8 feels like running Windows 7 with Windows WP7 on virtual machine.
I'll wait til MS can meld Metro and classic Windows together better.
Braumin - Thursday, March 1, 2012 - link
Yes, I have used it. I have used the Developer Preview, and am looking forward to the additons they have made in the CP for mouse and keyboard users.I think it looks great.
As with anything, you just need to get used to how it works. Soon it will be second nature to hit the corners with your mouse.
I am glad Classic Desktop is there - obviously the huge number of apps written for Windows need somewhere to run, but over time, you are going to use less and less of those as you replace them with Metro apps. Then you may never use the desktop again.
One thing I HATED in the DP was the multitasking. Only being able to have two apps open was a severe limitation at the time, and not being able to close apps you were not using made it terrible to switch apps, because there could be a large number of them to switch through. I think they changes they have made there with the task list, as well as the ability to drag windows off the desktop to close them, is a huge improvement.
Do I think it is going to be perfect? Of course not, but this is basically a Beta, and with the huge strides they have made in keyboard mouse usablitly from the DP to the CP, I only expect it to get better by the time the next version is out.