I'm not surprised
The IE machines didn't have to do much. The connected and then spent the entire 6 minutes displaying a pop up asking if you wanted to download a more modern browser.
Every one of us can do their bit to “save the planet”, whatever that means, and Microsoft has jumped aboard that bandwagon by commissioning a new study that investigates just how much electricity browsers consume when running on desktop and notebook computers. The Fraunhofer Center for Sustainable Energy Systems put its name …
Yep, a set time period seems like a thoroughly unfair test to me.
“We then measured the average power draw over one-second intervals for a six-minute period with each of the individual browsers open"
If I open IE on my work machine (which is the only time I open it), it huffs and puffs for a good while before it finally sorts itself out and loads the local intranet homepage.
A fairer test might be the total power draw over the time it takes to open the browser, load gmail, open and read a message, load YouTube, play one video and close again.
Never mind the power stats, I reckon IE on anything but a box fresh laptop would lose by about three minutes.
I have friends who run Linux distros on relatively old hardware ( coincidentally, they are the ones who asked me to install a good antivirus, Libre / Open office depending on how long ago it was, or just if I could make their boxes go a little faster ). They perform as well as newer hardware with Windows on it, but my guess is that they are less power hungry. Would Microsoft care to do a little research on that too?
It would also be interesting to see if there are variations when it comes to typing a text document or creating a spreadsheet...
Indeed… they should try booting up a Linux distribution, and see what Chrome and Firefox are like there compared to the Linux baseline.
Okay, we can't easily test what IE would be like under Linux since IE doesn't easily run under Linux (well, IE6 does... I've done it before and you can find instructions here), but it'd be interesting to see if the overheads of Firefox and Chrome running on Windows are anything like that on Linux.
Ohh, and if people are trying to reach the page in that screenshot, it's here.
When I ran that test, I found IE used more CPU time, while Firefox used more memory. Chrome didn't exist back then, in fact I think that test predates Webkit... although Webkit's daddy (KHTML; part of the Konqueror web browser) was lurking on that computer so I could've tried it at the time.
Yeah because W8 isn't more resource-efficient than W7, which was better than Vista. Oh wait, it is.
Yes because half the guts of IE aren't part of the OS and already running in the background, and Microsoft don't have a bunch of hidden API calls that other browser manufacturers can't reliably use.
Oh wait, they do.
And yet they can still only manage such slim margins over the browser with the renowned weight problem, Firefox? This... is not impressive.
that whole comparison is pretty pointless.
I can't help feeling that a better comparison would be between browsers running out of the box, and browsers run with the minimum of scripts (ideally none) and with the adverts blocked... it's hard to see how much power is required to display a static image...
Research published by a vendor, in a public journal, that shows the vendor's product is best? Be critical.
Research published by a vendor, in a press release, that shows the vendor's product is best? Be thoroughly skeptical.
Research published by a vendor, in a press release, that shows the vendor's product is best but doesn't describe the method used? Ignore it, as it probably came straight from the marketing department with no actually research being carried out.
You might want to look into the use of the hibernation feature. It is very useful.
Less than ten seconds from cold to working with all of your programs still running as you left them. Much less wear on your disk.
My computers both desktop and laptop hardly ever need to be booted any more.
How do people still not know about this and somehow not find out about it when they are moaning about boot-up times?
The technology has been supported since Windows 95.
Maybe, if you have a problem, rather than bleating moronically you could alternatively use the same general-purpose computing device you are sitting at or holding to perform a web search that may help you to solve the issue?
This post has been deleted by its author
Yes. 20 people were happy to 'me too' when a comment is MS-bashing. Not one reader thought to suggest the appropriate use of well-established power management features or to question that a four-minute boot time is by no means normal on modern hardware and might indicate some kind of more serious fault.
Your machine did not take that long to start when it was new. The company that wrote the operating system software installed on your computer are not responsible for the durability of your hardware. All of the cruft and possibly malware you have installed since then is not their fault. The onus falls on you as the owner to perform suitable periodic maintenance.
Hell, a good start might be to defrag the disc - there's a boot-time optimisation switch for just this purpose.
I'm pretty sure that you actually have a brain - how about putting it to some use? It's amazing what one can achieve when one puts his or her mind to solving problems, rather than always expecting someone else to change your nappy for you.
Are they really that stupid? Yes. That is why we call them "commentards".
Hmmm... First of all, how about running on Windows7 (or even XP)?
Second - I couldn't see what configurations they used for the various browsers. Were they running with plugins? If so, which? Did they represent a standard home setup?
Inquiring minds want to-... ah, who cares - it's an MS propaganda piece. Never trust a paid-for piece that says the paid-for item is better than the others. I'll wait for independent testing, TYVM.
The experiment only seems to measure power consumption in a static situation. Most browser use involves opening the browser, spending a period of time actually using the browser to achieve an aim and then closing it. The overall energy consumption is then significantly affected by how efficient the browser is - but that isn't tested.
So no : it's an interesting result, but not detailed enough to justify any meaningful conclusions.
I'm not sure why someone would open a browser, except if the power went off and they had to reboot. In any case, I just turned off JavaScript and the lights in the room got brighter, so you're probably right about the results not being detailed enough to justify any meaningful conclusions.
I'm the sort that doesn't leave things running in the background unless I actually want to use them in the near future, particularly browsers. Granted, if I'm in the minority then efficiency isn't necessarily significant.
I would have liked to have seen results based on representative operational scenarios, and these graphs sure aren't that.
It's all a bit academic, mind - (most) people choose browsers for what they do, not to be eco-friendly.
Hardly fruitful though.
Same goes for IE - it needs a more power hungry OS to host it - or at least I assume it does: W8 requires at least a 1Mhz processor. I've used firefox on a 90Mhz pentium so that suggests it can run on 1/10th the power. This is a valid comparison as the above test.
I have firefox/Adblock/flashblock and Internet Explorer on this corpo PC. The Internet Explorer "experience" is royal shit, because I can't block all those CPU-hungry flash ads, animated GIFs and all those other distractions. Just scrolling up and down with IE is a mess because it has to render some stupid commercial and it has to move that portion of the screen up and down.
Rendering movies and animated ads clearly is as power-hungry as it can get. So, factoing in Adblock and Flashblock, my guess is that firefox takes about 1/10th of energy consumption compared to Internet Explorer.
Regarding Fraunhofer, they are supposed to finance themselves by 70% from commercial funding and 30% from public funding. They had some massive successes in the past (e.g. MP3), but this smells of being the sales-bitch of M$. Nor much science to see here.
If they wanted to do real science, they should compare Flash player with VLC. I recently ran youtube videos properly using VLC, which would not run properly on the weak celeron Under Test using Flash Player.
During initial testing, we found that a variety of programs pre-installed by the computer manufacturers were resulting in significant fluctuations in computer power draw. Consequently, we removed all preinstalled software from all computers prior to testing and did a “clean” install of the operating system, Windows 8.
So the results are for computers maintained by a skilled Windows user, which is a tiny fraction of the installed base of home computers.
The dynamic benchmark test uses Microsoft's FishBowl benchmark, with 5 fish. Presumably this benchmark was optimised for IE, and not the others. The frame was not controlled. For all we know, Chrome and Firefox used more power because they had a higher frame rate.
(BTW, a Raspberry Pi uses at most 3.5Watts, and can update the FishBowl every 5 seconds at 1080p. Clearly Iceweasel on a Pi uses the least power ;-)
I doubt it has anything to do with undocumented APIs. Perhaps IE engineers have performance tuned their browser with respect to power consumption to look good in this study, or the other browsers are doing something which makes them look worse. I wouldn't leap to a conspiracy until such times as I knew the underlying reason.
Any web developer knows that to finish a web project, you need to get it to work in modern browsers, an then to try to get it to work in IE10, then in IE9 then in IE8, then in IE7 then in IE6 …
Not to mention that one reason we can’t all use HTML5 is because Certain Legacy Browsers can’t handle it.
To be honest, there's a lot of sites now that are made to give up and gracefully degrade as best they can while warning you that you're using a POS* browser.
Granted, IE still has its, erm, ways.
*Particularly Obsolete Software
It may be less power hungry on a client PC but what is the overall cost in terms of power?
How much power does the development office use compared to the others?
How much power do the development servers use compared to the other?
How much power is used in getting the developers to work?
Do the developers take pack up? Or do they use a canteen?
> Baseline: No browsers or other windows open
It is my understanding that Internet Explorer is so tightly integrated into Windows that parts of it are always running, which is often brought up in cold start comparisons between browsers. If this is the case then it would imply part of Internet Explorer's resource usage is included in the baseline. Perhaps further tests should be conducted with all three browsers open at the same time, and only one visiting a site at any given time?
>In addition, at the request of Microsoft we set the JavaScript timer frequency to “conserve power” in
the Windows power options. We found, however, that the default Javascript time frequency for all
computers tested was set to “maximum performance.” We did not investigate the impact of this setting
upon browser power draw.
I have to wonder if Firefox and Chrome use this system setting. It seems odd that they wouldn't run another set of tests at the default setting if nothing else.
1 - This was Microsoft commissioned. And Microsoft seem to have influenced the tests at a technical level. The test report specifically states:
a) ...We purchased 10 computers specified by Microsoft...
b) ...At the request of Microsoft we set the Javascript timer frequency to 'conserve power' in the Windows Power options....
One wonders why Microsoft were so anxious for THOSE computers to be used...?
2 - There is NO indication that the tests actually measured like-for-like. For instance, Chrome might actually be doing more work in caching and preparing the web site for better presentation than IE - which would naturally result in a slightly increased power draw.
3 - I use Opera, so I don't need to get involved in silly prick-measuring contests :)