Reality Check: Microsoft’s IE10 energy-efficiency claims

browser1

One of the official charts in the report, showing the difference in power consumption on notebook computers when running Chrome, Firefox and Microsoft’s Internet Explorer 10.

Microsoft’s Internet Explorer team has made big strides in recent years, supporting HTML5 and other web standards, collaborating on cool projects that show the potential of the web, betting big on touch input and generally attempting to make IE more cool.

But I had to chuckle this weekend as I dug into some new data released by Microsoft as part of its effort to get a leg up on Google Chrome and Mozilla Firefox in the browser market.

browsernew

At true scale: This chart shows the total difference in power consumption by notebooks when running Chrome, Firefox and IE10.

Internet Explorer 10 is “the most energy efficient browser on Windows 8,” the company declared — citing a report by the Center for Sustainable Energy Systems at Fraunhofer USA, which found that Internet Explorer uses up to 18 percent less energy than Chrome and Firefox.

“This means that if every Google Chrome and Mozilla Firefox user in the United States switched to Internet Explorer 10 on Windows 8 for a year, the energy saved could power over 10,000 households in the United States for that year,” the company said.

In other words: Use Internet Explorer 10, and help save the planet!

Well, sorta. Some additional context and thoughts about this report:

Microsoft paid the Center for Sustainable Energy Systems at Fraunhofer USA to conduct the study, the methods and results of which are summarized in this 21-page report.

10,000 households is 0.0087% of the 114.8 million households in the country.

Microsoft, of course, makes both Windows 8 and Internet Explorer. They’re part of the same division. Back in the day, the company argued they were basically the same product. The people who make IE know Windows inside and out. Of course IE10 is more energy-efficient on Windows 8. Do we need a study to find this out?

Some of the key charts in the report aren’t shown at true scale. Look again at the numbers on the y-axis in the chart at the top of this post, which shows average total power consumption of notebook computers, in watts, when running each of the three browsers across the 10 most-visited sites in the U.S. The numbers in that official chart stop at 15. I made this alternative version (above, right) to show the actual scale.

In other words, in the context of the overall power consumption of the computer, the differences in energy consumption among the various browsers is pretty darned small on most of the major sites tested in the study.

Yes, every little bit helps the environment, battery life is important, it’s great when companies can lead by example, and absolutely, we need to save the planet.

But if real people actually take this into account when choosing a web browser, well … more power to them, I guess.

  • Jrexxx

    would love to see what Safari looks like on OSX

  • Master Troll

    “10,000 households is 0.0087% of the 114.8 million households in the country.”
    Yes, it is called “a sample”. Duh… If the sample is well made and represents the universe of the study, it could even be only 1,000 households.

    • http://geekwire.com Todd Bishop

      Sorry, those numbers have nothing to do with sample size. Microsoft was making a statement to illustrate the potential impact of the energy savings if everyone were to switch from Chrome and Firefox to IE: It would be enough to power 10,000 homes for a year, the company says.

      I’m putting that into perspective by explaining that it translates into 0.0087% of households in the U.S.

      If you want to talk sample size, the report was based on testing of 10 computers (six notebooks and four desktops) running three browsers and accessing 10 websites.

  • Forrest Corbett

    I haven’t read the entire study, but after skimming through it I question their ability to gather accurate data. I see no mention of accounting for alternate code and/or functionality delivered based on user agent, nor different content being delivered because of any sort of dynamic content (eg. ads.) Also, they call the sites part of the test the”Static Website Test” – none of the websites listed are static sites.

    It’s also not immediately clear if they tested IE10 in metro mode or desktop mode. If they tested it in metro, there’s a good chance the Flash video test switched to HTML5. That would likely be more power efficient on its own.

  • http://www.techmansworld.com/ Michael Hazell

    I’m still sticking to Chrome.

  • Howard Jess

    It would be pretty difficult to make a real impact on US power consumption by trying to get users to change browsers. Most likely, though, there’s a lot more potential savings available, if Microsoft were to reduce the power required (by *all* browsers) to render the Bing site.

  • guest

    Take this meaningless study MS commissioned and multiple that out across the entire company and you begin to understand why their cost structure is so much higher than say Oracle. And the worst part is that right now some employee is putting this on their year end review and citing it as a major accomplishment.