backtop


Print 31 comment(s) - last by robinthakur.. on Jul 23 at 4:14 AM

Google didn't take action until Ian Morris' story gained traction

It seems as though Google issues a new public release of its Chrome browser every few weeks. While we’ve seen plenty of feature creep and performance improvements over the years, one Chrome for Windows bug has gone unpatched for four years.
 
The battery draining bug was first discovered by Chrome for Windows users way back in 2010 and is related to the “system clock tick rate” according to Ian Morris, a contribute to Forbes. Since the system clock tick rate in Windows can control how often a processor sleeps, it can have a noticeable impact on battery life.
 
Chrome for Windows sets the system clock tick rate to 1.000ms -- as long as it is open, no matter what the browser is doing -- instead of the customary 15.625ms. Microsoft says that setting the tick rate to 1.000ms can result in up to a 25 percent increase in power consumption, which explains the battery draining issues that Windows notebook users are experiencing when running Chrome.

 
Morris gives this real world example:
 
Well, when you open the most recent version of Internet Explorer, the rate stays at 15.625ms until the browser needs to do something where the rate must increase. If you go to YouTube, say, and play a video IE will increase the rate to 1.00ms. When you shut that tab, and carry on with normal browsing, it will return to 15.625ms. In Chrome though, it is increasing the rate as soon as the browser is opened, and it keeps it high until you shut the browser completely.
 
Following Morris’ article on the subject, Google has finally acknowledged the issue and is working internally to found a solution.

Sources: Forbes via, PC World



Comments     Threshold


This article is over a month old, voting and posting comments is disabled

This bug does not exist
By Labotomizer on 7/18/2014 10:39:15 AM , Rating: 2
It's a Google product. This cannot be true and is nothing but propaganda against the benevolent internet company. stop lying!




RE: This bug does not exist
By retrospooty on 7/18/2014 11:03:07 AM , Rating: 1
That is a very odd chip you seem to have on your shoulder there. Maybe you should just use other products that dont hurt your butt.


RE: This bug does not exist
By Labotomizer on 7/18/2014 11:54:09 AM , Rating: 3
Hmmm... or maybe I was just making a joke. It's software, I don't expect any software to be perfect. I've never noticed any difference in battery life when using Chrome on my laptop or Surface. Certainly not butt hurt.


RE: This bug does not exist
By Reclaimer77 on 7/18/14, Rating: 0
RE: This bug does not exist
By Labotomizer on 7/18/2014 12:18:50 PM , Rating: 2
No, you're right. It would have better fit an Apple article. Users who think usually understand there's a reason things get patched.


RE: This bug does not exist
By inighthawki on 7/18/2014 1:06:19 PM , Rating: 1
Google does have a pretty good track record of having fairly stable products, though. I'll give them that :)


RE: This bug does not exist
By StevoLincolnite on 7/18/2014 6:36:21 PM , Rating: 2
Ironically enough, that never used to be the case, all their software/web products were always in Beta for what seemed like years!

Despite that, they were still always incredibly usable.


RE: This bug does not exist
By inighthawki on 7/20/2014 2:28:10 AM , Rating: 2
But that's what I mean. Their products, even during beta, were often pretty stable... I wasn't actually implying anything about beta vs release.


RE: This bug does not exist
By Reclaimer77 on 7/18/14, Rating: 0
RE: This bug does not exist
By amanojaku on 7/18/2014 11:12:23 AM , Rating: 2
quote:
Who not only denies bugs and viruses exist on their pristine ecosystem, but convinces millions of it's cult following that this is the case as well.
Yup, pretty much the reason I sh*t on the company any time I can. It's not like Apple makes inferior products, but they certainly aren't superior, either. Google isn't perfect but it's far more humble, so it's easier to forgive its mistakes.


RE: This bug does not exist
By Labotomizer on 7/18/2014 11:58:31 AM , Rating: 2
Oh come now, when you make a joke I don't call it trolling when it's at the expense of a platform I like. You know full well this was sarcasm.


RE: This bug does not exist
By inighthawki on 7/18/2014 12:11:38 PM , Rating: 2
Yep. Last I checked, sarcasm and jokes don't fall under the category of trolling :)


RE: This bug does not exist
By Reclaimer77 on 7/18/14, Rating: 0
RE: This bug does not exist
By Labotomizer on 7/18/2014 12:17:45 PM , Rating: 2
I didn't even mean it slanted at Google honestly. It just happened to be a Google article. When people blindly defend a company or, just as bad, bash a company for no other reason than it being said company I find it irritating. Could have been Apple or MS. I'd include BlackBerry but there's no one who defends them anymore...


RE: This bug does not exist
By Reclaimer77 on 7/18/2014 12:31:26 PM , Rating: 2
I must be still hungover, because something isn't making sense lol.

You were using sarcasm, in the first person perspective writing style. As a reader, that seemed to be you indicating people really DID say and think things like that about Google?

I feel like I'm missing something.....lol ah well, screw it. Who cares? Good joke, and I see what you mean NOW. I just didn't see it then.

It's probably just me...you know those caffeinated water additives you are supposed to squirt in a glass of water? Yeah I've sorta been just straight squirting them in my mouth to try and get me through the day.

my kingdom for the weekend....weekend please come


RE: This bug does not exist
By Labotomizer on 7/18/2014 1:56:53 PM , Rating: 2
It's okay, I got woke up at 2 AM and have been working off two hours of sleep... So it was likely "infinitely" funnier in my head than it came across.


RE: This bug does not exist
By HolgerDK on 7/18/2014 11:23:48 AM , Rating: 2
In these comments, people who didnt see the sarcasm in the first comment...


RE: This bug does not exist
By amanojaku on 7/18/14, Rating: -1
Finally
By cditty on 7/18/2014 11:33:00 AM , Rating: 3
I'm glad they are fixing this... I use my laptop on battery a lot and have been following this bug for a long time. I honestly thought they would never fix it because of their stance with Microsoft. As a user of multiple platforms, I use Chrome for it's excellent cross platform syncing.

I do wish they would stop dragging their feet on fixing issues when it adversely affects Microsoft products. I understand the reasoning, but they are affecting their users, not just Microsoft. Having an Android phone & tablet plus a Windows PC or laptop is common.




RE: Finally
By Reclaimer77 on 7/18/2014 11:42:56 AM , Rating: 1
It's just a bug, not a conspiracy. Google makes a massive amount of money off of people using Windows. The theory that they would WANT to give them a bad experience, as if that hurts Microsoft, is the worst kind of speculative thought.


RE: Finally
By Labotomizer on 7/18/2014 11:57:06 AM , Rating: 1
I also think the article is misleading. A 25% power increase from system clock certainly isn't overall power increase. And considering that we're talking about a web browser not requiring a great deal of compute power most of the time we're talking about a 25% increase for the CPU at a low power state, and that's likely just theoretical number anyway. On an 8 hour battery we might be talking 15-20 minutes of extra drain and I would wager that's on the high side.


RE: Finally
By inighthawki on 7/18/2014 12:27:02 PM , Rating: 2
quote:
A 25% power increase from system clock certainly isn't overall power increase.

It's not a "25% increase in power usage by the system clock," it's "25% total power usage increase"

Running the system clock at 1ms over 16ms can really drain the battery quite significantly by waking up the CPU cores far more often, preventing them from entering low power states.


RE: Finally
By Labotomizer on 7/18/2014 1:58:36 PM , Rating: 1
It could only mean CPU. Screen is the biggest power draw so there's no way the system clock increases screen draw by 25%.


RE: Finally
By Omoronovo on 7/18/2014 4:04:04 PM , Rating: 3
You are only correct about the display using the most amount of power if the cpu is able to power gate and idle properly.

A clock tick rate of 1ms versus 15.625ms means that the cpu has less than 1ms of sleep time between timer calls (at most), before the system wakes it up. As anyone who understands the topic will tell you, cpu timers are one of the primary causes for preventing modern cpu architectures from properly activating their power gating techniques; we're talking the difference between the cpu having sections that are physically off (sections of cache for example) versus being powered up and idling. If the cpu is being woken up 15.625 times more often than is needed, then it is definitely reasonable that battery life can be shortened considerably.

For more information (instead of being spoon fed), see the following sources: http://blogs.technet.com/b/askperf/archive/2009/10... for a short overview of the principle behind this change (made in windows 7 onwards), http://www.cs.utah.edu/~niti/hpca11.pdf for a research paper from the computing school of the university of Utah which explains why power gating is so important, and http://download.microsoft.com/download/9/C/5/9C5B2... for an (in-depth) implementation white paper from Microsoft on how much this can affect performance and energy consumption.

In addition, Microsoft introduced the idea of the "tickless kernel", something which linux has been capable of for a while, and uses a hybrid mode of timer polling from windows 8 onwards. This means the timer can be higher than 15.625 if applications are written to enable Windows to coalesce their timers. Chrome has not been, so on windows 8 the battery drain could be even more severe, though it depends what else is running at the time. The fastest timer is always used, and if that happens to be the 1ms of Chrome, then that's what's used, regardless of its requirement.

I hope this helps your understanding of the issue.


RE: Finally
By Omoronovo on 7/18/2014 4:25:01 PM , Rating: 3
Regarding my display comment, I want to clarify:

The display of most modern notebooks can use anywhere between 40 and 60% of the power at any given moment; incredibly variable of course between size, resolution, brightness, and technology. This leaves between 40 and 60% for the rest of the machine; a 25% reduction in battery life (after taking display out of the equation) is very easy to accomplish by simply preventing the cpu from idling (at which point it would go from ~1-3% usage to ~8%, a 2-3x increase in direct draw but a considerably larger 15-25% increase in the drain of the remaining battery charge (which is what is directly measured as battery life in use).

Hopefully that makes sense.


RE: Finally
By inighthawki on 7/18/2014 12:10:28 PM , Rating: 2
It's probably just lower priority than other bugs and features. Although it's not uncommon for software companies to focus their priorities on their own platforms, I think you are right in that Chrome on Windows is a big enough deal that they focus on it. But it just sounds like this is a slightly uncommon bug (I don't know of anyone who has complained about this issue).


RE: Finally
By Labotomizer on 7/18/2014 12:20:15 PM , Rating: 1
Chrome on Windows is very important to Google. Chrome on Metro isn't but I don't know of anyone who really cares about that anyway.


RE: Finally
By Reclaimer77 on 7/18/2014 12:24:44 PM , Rating: 1
This article never even bothers to quantify what kind of power drain we're talking here. It makes it sound like Chrome is just straight up hammering battery life, but that hardly seems like that's the case.

Like Labotomizer said, I think this is made to appear to be a bigger issue than it is. There's no way Chrome is causing a 25% loss in battery life.

I think we would have heard about this way before now if it was that significant. I haven't, and neither have you. And for whatever it's worth, I know people who use Chrome on laptops and they never mentioned anything about a battery issue.

Oh and just so a few ppl's heads can explode, I use Firefox on my laptop. Never really got into Chrome myself, prefer Firefox's feature set and out-of-the box tab management.


RE: Finally
By inighthawki on 7/18/2014 12:29:28 PM , Rating: 2
Raising the system clock frequency is a pretty big power hog. It can prevent CPU cores from going idle or into low power states. There's a good reason that the system clock typically runs ate 16ms instead of 1ms. I would not at all be shocked if the total power usage increased 25% as a result.


RE: Finally
By frostyfiredude on 7/20/2014 12:30:55 AM , Rating: 2
I have a Surface Pro 2, so a Haswell based device. Chrome versus IE11 has a very significant effect on my usage. Getting 11 to 12 hours of battery life with a light usage pattern such as browsing Reddit at ~50% brightness is common with IE11 where a similar usage with Chrome sees 8-9 hours. How much of that is the timer issue is beyond me, but Chrome in general has a resource usage issue.


RE: Finally
By robinthakur on 7/23/2014 4:14:50 AM , Rating: 2
I find it hard to believe that a battery draining bug that only affects windows on one of the most used apps has been around for years without being fixed and that this is not somehow part of Google's civil disobedience against MS, much like not making a Youtube app for Windows Phone or Windows apps for Windows 8. Chrome on the Mac is not listed as affected, possibly because the battery life on Mac is highly publicised and would loudly shout if they didn't meet expectations. Conversely people expect the battery life on a pc laptop to be dodgy at the best of times.


"What would I do? I'd shut it down and give the money back to the shareholders." -- Michael Dell, after being asked what to do with Apple Computer in 1997











botimage
Copyright 2014 DailyTech LLC. - RSS Feed | Advertise | About Us | Ethics | FAQ | Terms, Conditions & Privacy Information | Kristopher Kubicki