Slashdot Mirror

← Back to Stories (view on slashdot.org)

Google Chrome Bests Microsoft Edge, Mozilla Firefox, Opera In Independent Battery Life Tests (betanews.com)

Posted by msmash on from the getting-to-the-bottom-of-things dept.

An anonymous reader shares a report: YouTuber Linus Tech Tips has pitted Microsoft Edge against Google Chrome, Mozilla Firefox and Opera and discovered that it does not deliver as strong a performance as Microsoft claims. Linus Tech Tips took four Dell Inspiron laptops, with the same specs, and found that Microsoft Edge trails Chrome and Opera in battery life tests. It would seem that it still beats Firefox, after all. However, the results are much, much closer than what Microsoft's own tests indicate. On average, the difference between Chrome, which offers the best battery life, and Microsoft Edge is under 40 minutes. Opera comes closer to Microsoft Edge than Chrome in this test. Even Creators Update, which based on Microsoft's test should help Microsoft Edge obliterate the competition, didn't help make it faster than Chrome. Linus says he used the same methodology that Microsoft used in its set of battery tests earlier this year, in which it declared Edge as the winner.

64 of 114 comments (clear)

  1. Manufacturers lie about their products... by gweihir · · Score: 4, Insightful

    What else is new. In particular, about everything MS praises in Windows is either bogus or actually a disadvantage.

    --
    Most ACs are not even worth the keystrokes to insult them. Be generically insulted by this and ignored otherwise.
    1. Re:Manufacturers lie about their products... by courteaudotbiz · · Score: 1

      But marketing is king, and being able to make people believe your product is better has nothing to do with your product being actually better.

  2. And nobody is surprised.... by tomxor · · Score: 1

    How many times has Microsoft made false claims about it's browser, all their IE/Edge press BS originates from contrived statistics or benchmarks. You can never trust them and they will never change.

    1. Re:And nobody is surprised.... by Gadget_Guy · · Score: 1

      It's unfair to say that the original claims were false based on the new tests. The new tests ran different versions of the browsers. Since Microsoft started bleating about battery life, the other browsers started paying attention to their power usage. Chrome has probably doubled its version number since MS first spoke about it. OK not really, but they have had enough time to close the gap (assuming one really existed).

      The new test also ran on different hardware than Microsoft used. Most importantly, the Inspiron 15 7559 laptop uses an NVIDIA GeForce GTX 960M instead of Intel's chipset. If for some reason some video acceleration doesn't work on the GeForce, then that could indeed level the playing field as all browsers would be using the CPU to play the video and graphic work. (The opposite could be true too where the GeForce does a better job at accelerating all browsers more consistently than Intel did.)

      All this isn't to say that some funny business didn't go on in Microsoft's test, but it seems less likely considering that they did publish their test procedure so that it could be duplicated by others. That said, it would not surprise me if there is such a wide variation in each run that Microsoft didn't choose the best one that supports their browser.

      My personal opinion is that Edge does indeed result in lower battery usage, simply because it offers such a poor user experience that it makes you want to close the browser sooner than the other brands.

    2. Re:And nobody is surprised.... by supremebob · · Score: 1

      It is at least somewhat surprising, considering that Microsoft has been advertising this "fact" in the Windows 10 Notification Center for months if you happen to be a Chrome user.

      Why would Microsoft tell this lie to millions of people using Chrome in Windows 10, knowing that at least a small percentage of them have to means to prove that it's BS?

    3. Re:And nobody is surprised.... by tomxor · · Score: 1

      Perhaps isn't not as wide spread as before but they have publically anounced similar performance and security related comparisons putting IEDGE on top, they have all been debunked - I suppose the difference is that was probably only towards a mostly developer community.

    4. Re:And nobody is surprised.... by tomxor · · Score: 1

      All this isn't to say that some funny business didn't go on in Microsoft's test, but it seems less likely considering that they did publish their test procedure so that it could be duplicated by others. That said, it would not surprise me if there is such a wide variation in each run that Microsoft didn't choose the best one that supports their browser.

      dude... that's the point, I'm not giving MS benefit of the doubt, they repeatedly cheated in these things, you give them way too much credit by suggesting others are more likely to purposely disadvantage them in a test. MS will do contrive an unfair test at the drop of a hat... every single time.

    5. Re:And nobody is surprised.... by Gadget_Guy · · Score: 1

      And yet they published everything that you needed to use to reproduce the tests and none of the other browser manufacturers demonstrated that it was wrong. They did however improve their battery usage in subsequent versions. Do you think that the browser makers would go to that trouble instead of taking the easier path and just producing their own results?

      If you want to say they have cheated then show the evidence. And no, doing the tests of browsers that are four versions older than Microsoft used is not good enough.

  3. In other news by DontBeAMoran · · Score: 1

    Safari beats Edge on the latest versions of macOS.

    --
    #DeleteFacebook
    1. Re:In other news by TheFakeTimCook · · Score: 1

      Safari beats Edge on the latest versions of macOS.

      And I'm DAMN sure it beats resource-hogging Chrome.

    2. Re:In other news by TheFakeTimCook · · Score: 1

      Sorry to reply to my own post; but here is a citation for Safari beats Chrome, which is just as relevant as Chrome beats Edge; since both Safari and Edge are single-platform browsers:

      See, esp. Footnote 2.

      https://daringfireball.net/201...

    3. Re: In other news by DontBeAMoran · · Score: 1

      Who said anything about iOS? My post clearly said macOS.

      --
      #DeleteFacebook
    4. Re: In other news by Bert64 · · Score: 1

      Most embedded systems have only one rendering engine, it would be a waste of resources to have several running at once and a lot of extra work to make the api modular so it could be swapped out (not to mention the inevitable incompatibilities if the replacement engines differed slightly).

      --
      http://spamdecoy.net - free throwaway anonymous email - avoid spam!
    5. Re:In other news by fellip_nectar · · Score: 1

      Well, who wouldn't be afraid of browser competition when their browser sucks ass?

      --
      Worst. Signature. Ever.
    6. Re: In other news by lars.jeppesen · · Score: 1

      You are absolutely correct. Small correction: Google forked webkit long time ago and is using Blink as browser engine (along with Opera). But on iOS it's not allowed of course

  4. Well well well. by Dust038 · · Score: 1

    Something Something, Car Emission Reports...

  5. Testing methodology by TWX · · Score: 4, Insightful

    So did the reviewer, upon completing the first round with the four machines, then rotate the software under-test across the machines, rerun, rotate again, rerun, etc?

    What were the parameters of the test? Was this some kind of scripting that compelled the browser to pull content without user interaction? How was that achieved, and could extra usage from that software have skewed results? What content was pulled-down? Were different kinds of content, reflecting different kinds of users/usage pulled-down?

    I ask all of this because it affects the results. A single browser on a single laptop is a sample size of one. If the testing involved four out-of-the-box laptops with new batteries an dfour browsers, then one has a single data point for each browser. More testing is probably necessary to establish real results instead of just generating fanboy arguments.

    --
    Do not look into laser with remaining eye.
    1. Re:Testing methodology by __aaclcg7560 · · Score: 2

      This is Linus Tech Tips. It's entertainment. I wouldn't take it too seriously.

    2. Re:Testing methodology by Anonymous Coward · · Score: 1

      If you watch the video, he uses 4 laptops. Does a base run to get battery life for all of them. Then does one browser on all four, then does each browser on each laptop, multiple times, before and after the latest windows 10 'service pack'. Seems pretty well done to me. YMMV

    3. Re:Testing methodology by Anonymous Coward · · Score: 1, Informative

      Maybe, but Linus did everything that GP asked for.

      Meanwhile, Microsoft didn't in their own tests, and in fact used just a sample size of one. If anything, I wouldn't take Microsoft's numbers seriously.

    4. Re:Testing methodology by AmiMoJo · · Score: 1

      Didn't watch the video, but TFA mentions a 40 minutes difference which strongly indicates that they don't know what they are doing. 40 minutes is meaningless, maybe Edge lasted 1 minute and Chrome lasted 41 minutes. If they had a clue they would have expressed it as a percentage.

      --
      const int one = 65536; (Silvermoon, Texture.cs)
      SJW, n: "Someone I don't like, and by the way I'm a fuckwit" - AC
    5. Re:Testing methodology by wbo · · Score: 1

      I wonder how the power profile settings were configured on the machines. Edge and IE altar their behavior quite significantly depending on how those settings are configured. Last I checked, Chrome outright ignores most of those settings while Firefox at least attempts to respect some of them.

      Settings such as minimum Javascript timer frequency, GPU power profile (impacts GPU compositing) can have a big impact on battery life.

      Also important is whether or not the power profile was set to throttle the CPU when on battery power and how much to throttle. (Setting it up so that the system is passively cooled while on battery via CPU and GPU throttling can greatly increase battery life at the expense of peak performance.)

      Generally while web browsing you don't need 100% CPU and GPU power - especially when operating off of a battery so throttling makes sense. However if the power plan was set to tell applications that it is OK to use 100% of CPU and GPU resources all of the time while on battery power then applications typically won't try to reduce usage and battery life will be significantly worse as a result.

    6. Re:Testing methodology by gweihir · · Score: 1

      Fulfills scientific standards IMO, as long as absolute accuracy is not a factor. For relative measurements, this is entirely fine.

      --
      Most ACs are not even worth the keystrokes to insult them. Be generically insulted by this and ignored otherwise.
    7. Re:Testing methodology by TWX · · Score: 1

      Nice try, given that I'm typing this on a Debian box in FF.

      I asked what I asked because if I'm going to rely on someone's test results I want to be sure that the tests performed were thoughtfully designed and minimize the potential for the results to be skewed due to unaccounted-for issues in the testing procedures. Having done QA software testing myself, I know that it can be a challenge to develop realistic tests that actually present the bulk of use-cases for software to get valid results. Any fool can create a testing regimen that presents what they want to see, but it takes work to create tests that reflect reality closely enough to be useful and to make it hard for the unscrupulous to manipulate testing conditions.

      If you want proof that a vendor would do that, look at all of the car company scandals involving cheating on emissions results. Prior to the 1996 model year, after cars got to six or seven years old they had to be tailpipe-tested in many places, so car companies were not free to cheat using computer manipulation because the cars, still within their undisclosed emissions warranties, were going to be retested while the manufacturer was on the hook for problems. 1996 an thereafter the automaker gets to tell the emissions test station that the car passes, rather than the station performing its own independent tests, so car companies that use trickery to pass at the government certification centers and emit whatever they want elsewhere can report that the car does indeed pass even if in actuality it doesn't.

      --
      Do not look into laser with remaining eye.
    8. Re:Testing methodology by TWX · · Score: 1

      Having watched a couple of Linus Tech Tips videos in the past and not been terribly impressed I was not expecting much in the way of good procedures. Since I can't watch the video right now, better to ask the questions and get a discussion on the nature of software testing going.

      If he/they have managed to improve the quality of the youtube channel, good for them.

      --
      Do not look into laser with remaining eye.
    9. Re:Testing methodology by Walking+The+Walk · · Score: 1

      So did the reviewer, upon completing the first round with the four machines, then rotate the software under-test across the machines, rerun, rotate again, rerun, etc?

      I watched the video, and yes, they rotated the tests. They first ran multiple baselines on each machine to identify the variances and weed out any dud machines. Then they ran the test multiple times on each machine, using each of the four browsers multiple times. This testing took a long, long time, because they had to wait for all of the laptops' batteries to run dry before ending each test, then fully charge them before the next test.

      What were the parameters of the test? Was this some kind of scripting that compelled the browser to pull content without user interaction? How was that achieved, and could extra usage from that software have skewed results? What content was pulled-down? Were different kinds of content, reflecting different kinds of users/usage pulled-down?

      He explains the test parameters within the first few minutes of the video. They ran their own tests first, to measure how the browsers would respond to "normal" usage (some browsing and then watching a video until the battery dies.) Then they followed Microsoft's published test methodology, using all four machines and running each browser on each of the machines in multiple tests. Then they show the minimum, average and maximum for each browser, and include machine-specific results so you can see the variance across machines. They also tested with and without the recent Windows 10 Creators update, which to me gave the most surprising results. (Spoiler: before the update there was huge variance in Edge results, even on the same machine! After the update the variance was very small.)

      I ask all of this because it affects the results. A single browser on a single laptop is a sample size of one. If the testing involved four out-of-the-box laptops with new batteries an dfour browsers, then one has a single data point for each browser. More testing is probably necessary to establish real results instead of just generating fanboy arguments.

      If you're going to say that someone should follow proper scientific method, please actually check first to see if they did. It's not like this was a 50 page paper on ArXiv, this is just a shortish Youtube video.

      --
      A recursive sig
      Can impart wisdom and truth
      Call proc signature()
  6. Re:Is the test system fair? by Njorthbiatr · · Score: 1

    You could, you know, watch the video where they describe it.

  7. No stupid videos? by samwichse · · Score: 1

    Can someone just post a table of the results they got. Not watching a stupid video for information I could absorb in 15 seconds from text.

    1. Re:No stupid videos? by Anonymous Coward · · Score: 1

      No. They can put their ads (which I will ignore) on the page with the table of results. They can get their ad revenue the same way other web pages do. The video could be interesting background material, but who wants to watch a long video when a simple table of results will convey the info?

    2. Re:No stupid videos? by samwichse · · Score: 1

      I feel like there's some huge logical disconnect in your mind here.

  8. How did Internet Explorer do? by jfdavis668 · · Score: 1

    Since we are stuck using it for many of our systems, it would be helpful to know how it rates.

  9. Serious Edge... by emil · · Score: 1

    ...is open source Edge.

    Microsoft, you must abandon hope that you will regain any market share with a closed source browser locked to Windows 10. It will never, never happen.

    When independent ports of Edge emerge on OSX, Ubuntu, Dragonfly, and Haiku, then perhaps there will be hope - and not one day before.

    1. Re:Serious Edge... by dabadab · · Score: 1

      If you trust the quality and security of MS products and the business practices of the company enough to endorse them by using their products (or if you're too damned oblivious to have put a moment's thought into it, American-style) then you're already using Windows 10 and maybe Edge as well.

      No, not really. Just look at Visual Studio Code: it is pretty popular with web developers across all platforms - even I have found myself running it on my Debian PC. And I don't use it to "endorse them" but because it's the best tool I have found for the job.

      --
      Real life is overrated.
  10. The Win10 music player sucks too? by emil · · Score: 1

    Tell me why I need Windows 10 again?

    http://www.tedunangst.com/flak/post/efficient-music-players-remain-elusive

    1. Re:The Win10 music player sucks too? by thegarbz · · Score: 1

      You don't. Until other versions so being supported anyway.

  11. Re:Laptops - heh. by Hentes · · Score: 1

    The aim of the test was to verify/dispel Microsoft's claims that browsing on Edge will make your battery last longer. Edge doesn't have an Android version, and the market share of Windows Phones is too small to be relevant.

  12. Irrelevant by Aethedor · · Score: 4, Insightful

    It's a web browser. It gives me access to the world wide web. Security and privacy are far, far more important than battery life. Chrome is made by Google, which is an absolute no-go when it comes to privacy. So, thanks, but no Chrome for me.

    --
    It doesn't have to be like this. All we need to do is make sure we keep talking.
    1. Re:Irrelevant by oakgrove · · Score: 3, Interesting

      You don't think MS' telemetry and browser is gathering information on you at 10 times the rate of Google Chrome? If you cared that much, you'd be using an OS that doesn't support Edge in the first place.

      --
      The soylentnews experiment has been a dismal failure.
    2. Re:Irrelevant by Aethedor · · Score: 1

      Who says I'm using a Microsoft OS while browsing the internet?

      --
      It doesn't have to be like this. All we need to do is make sure we keep talking.
    3. Re:Irrelevant by Aethedor · · Score: 1

      :D Dude, chill.... But if you're really that frustrated, I'll help you relax by telling what I use: Firefox on both Linux and MacOS. You can sit down now and have some coffee. :)

      --
      It doesn't have to be like this. All we need to do is make sure we keep talking.
    4. Re:Irrelevant by Bert64 · · Score: 1

      At least you have choices...
      Chromium is open source, and de-googled versions are available. And there's always firefox.
      Apple also don't seem to be so interested in collecting user data as MS and google, and there's also opera.

      --
      http://spamdecoy.net - free throwaway anonymous email - avoid spam!
  13. iOS could never be an option. by emil · · Score: 1

    Apple does not allow browsers on that platform that do anything other than wrap Webview.

  14. Configuration by Anonymous Coward · · Score: 2, Informative

    A family member's machine was "running slow" so I setup Firefox and configured it not to use disk cache at all.
    She has broadband internet and does not go near her monthly limit and frankly I think the network request can serve the content up faster than the laptop hard drive.
    Google Chrome and other browsers lack the feature to really minimize disk writes by eliminating disk cache.
    RAM is cheap and she generally just puts the thing in standby, meaning sites are still cached in RAM if their cache metadata hasn't expired them.

    I'll bet it would beat all of the tested configurations hands down for battery life (energy efficiency and battery lifetime) as well as storage device (HDD, etc.) lifetime.

    Disk cache doesn't make sense, so if Firefox is the only browser that TRULY does not write to disk, then reward them for their forward thinking if their browser has configuration settings that can tilt the axis of the test.

    1. Re:Configuration by Bert64 · · Score: 1

      As with all things it depends on use cases, i'm often on slow and/or metered connections but just as often on fast connections, so disk caching often does make sense. Plus i often run other things than just a browser, so i don't want all my ram occupied by cache.

      --
      http://spamdecoy.net - free throwaway anonymous email - avoid spam!
  15. Microsoft code by emil · · Score: 1

    If you trust the quality and security of MS products and the business practices of the company enough to endorse them by using their products (or if you're too damned oblivious to have put a moment's thought into it, American-style) then you're already using Windows 10 and maybe Edge as well.

    There are some coders at Microsoft who are the very best in the business. If Microsoft releases their code for practical applications, why would we discriminate against it merely for the name of the originator?

    Before Dave Cutler designed the kernel for Windows NT, he had previously lead the VAX VMS kernel design team. He produces tight and beautiful code, and ignoring any open-source practical contributions from him would be quite foolish.

    That said, it's unlikely that Cutler has anything to do with Win10, and it shows.

    1. Re:Microsoft code by phantomfive · · Score: 1

      Before Dave Cutler [wikipedia.org] designed the kernel for Windows NT, he had previously lead the VAX VMS kernel design team. He produces tight and beautiful code,

      Have you seen his code? Is it available somewhere?

      --
      "First they came for the slanderers and i said nothing."
    2. Re:Microsoft code by __aaclcg7560 · · Score: 1

      Before Dave Cutler designed the kernel for Windows NT, he had previously lead the VAX VMS kernel design team.

      The story behind Dave Cutler and Windows NT was covered in "Showstopper!: The Breakneck Race to Create Windows NT and the Next Generation at Microsoft" by G. Pascal Zachary. A good read. WinNT was developed the same time that Microsoft was still playing nice with IBM and OS/2.

  16. Re:what about... by nctritech · · Score: 1

    Adblock Plus is available for IE, Edge, Firefox, Chrome, Safari, Opera, and a few other browsers, as well as any browser (Pale Moon, Chromium, etc.) based on the previously mentioned browsers.

  17. How to Ctrl+F a video? by tepples · · Score: 1

    Watch the video

    How would I Ctrl+F a video or watch it on a severely slow and/or capped Internet connection? And how would I watch without audio on my computer or with a hearing disability? Its captions have errors, such as a complete lack of punctuation and goofs like "until seventh generation core i7" when "Intel" was meant, due to having been automatically generated. Pending answers to these, I prefer text over video.

    1. Re:How to Ctrl+F a video? by Bert64 · · Score: 1

      Yeah i have the same problem these days, i cant stand the prevalence of videos especially ones which provide instructions on technical subjects... I'm often on a slow or metered connection, and often the instructions are based around something textual (eg typing commands) which i would much prefer to cut+paste than try to read/listen from a video... The video may be compressed heavily (only way to make it playable on a poor connection) making text unreadable, and the accent might be hard to understand. Textual commands have no misunderstanding, and no human error if you cut+paste.

      --
      http://spamdecoy.net - free throwaway anonymous email - avoid spam!
  18. How to search video captions? by tepples · · Score: 1

    I would, but Firefox's "Find in page" (Ctrl+F) in a YouTube video page searches only the comments, not the video captions. What am I missing?

  19. just a feeling by epine · · Score: 2

    Just a feeling, but I suspect HTML 2.0 would have lapped the field, and then gone out for an late-night pub crawl, returning home at the cock of dawn to romp nekid until the sun crosses over the yard arm with an insatiable beer-goggles Wonder Bra, and still find enough energy in the tank to chew off his arm a few hours later.

  20. Opera Mini works around it by tepples · · Score: 1

    Opera Mini renders HTML documents remotely in order to work around the Apple WebKit restriction. In addition, different wrappers for Apple WebKit can have different battery life characteristics, though I imagine the difference is negligible. Among Opera Mini rendering remotely, Safari wrapping Apple WebKit, Chrome wrapping Apple WebKit, and Firefox wrapping Apple WebKit, which of the four can browse for the longest on a single charge?

  21. RAM is cheap until you've already maxed your PC by tepples · · Score: 1

    RAM is cheap

    Unless you've already installed the largest RAM modules that your PC will take. I've seen new PCs for sale with no option to reconfigure for more RAM. For example, this product detail page for an Inspiron 11 3000 laptop mentions "2GB" but doesn't offer any choice to upgrade at build time nor state what sort of RAM slot it has.

  22. Do not look at this code... by emil · · Score: 1

    ...if you are a kernel or systems contributor to a major platform.

    As I remember, there were compromises in the kernel architecture that were evoked deep "regret" in the commentary.

    1. Re: Do not look at this code... by Bert64 · · Score: 1

      Not in cases where MS actually pays people to use their products (which has happened in quite a few places in order to prevent high profile customers moving to linux)

      --
      http://spamdecoy.net - free throwaway anonymous email - avoid spam!
  23. Edge Mini anyone? by tepples · · Score: 1

    Apple allows only one browser engine on iOS and that engine is Webkit.

    The other option is remote rendering, as used by Opera Mini. Microsoft could likewise run EdgeHTML in its huge Azure cloud and send the result of remote rendering to iOS devices.

  24. Not surprising by OneHundredAndTen · · Score: 1

    The upshot is, Microsoft, once again, has been caught lying. Which is, of course, not news. Consider yourself middle-fingered, Microsoft.

  25. What Version of FireFox was used? by Anonymous Coward · · Score: 1

    The real question that we should be asking is what version of FireFox did he use? v54 just debuted with better RAM management, wouldn't that directly translate to improved battery life?

    1. Re:What Version of FireFox was used? by viperidaenz · · Score: 1

      The amount of data stored in RAM doesn't impact power consumption.
      If Firefox is spending more CPU time to better manage RAM usage, it may equate to worse battery life.

  26. It's good news by viperidaenz · · Score: 3, Funny

    Windows has recently stopped reminding me that switching to Edge will gain me 2 hours battery life, despite my laptop being plugged in 95% of the time.
    They're now telling me constantly that I need to adjust my screen brightness settings to save my battery, despite being plugged in to an external monitor with no backlight adjustment capability and hence no control to change it in the settings this "helpful" tip takes me to.

    Go Windows 10!

  27. Re:Laptops - heh. by ChunderDownunder · · Score: 1

    Sob... I and the other 2 users of Firefox OS will tell you Mozilla was ahead of the curve but sadly killed the platform.

  28. Battery Saver mode by tezbobobo · · Score: 1

    Does anyone know if Opera was used in battery saver mode?

  29. Duck it by jf_moreira · · Score: 1

    Duck battery, I need a browser who actually works, and Chrome is NOT one that work with all types of websites. Frequently I have to launch Firefox for that and always keep it installed when using Chrome. That's why, up to today, I never totally migrated into Chrome. I still use Chrome when youtubing because of the extensions, and that is the only reason. At work I never use Chrome.