Looking for Honesty in Battery Life
Posted by: Stephen Wildstrom on June 18, 2009
When you are a distant No. 2, you really have to try harder. And sometimes it even works.
A few years ago, AMD launched a campaign against using a processor’s clock speed as the principal metric of computer performance. After push speeds to the point where computers became more notable for their heat production and energy consumption than their computational speed, Intel had to concede the point.
Now AMD has taken up arms against a widely used and thoroughly misleading standard of laptop battery life called MobileMark 2007. Patrick Moorhead, AMD marketing vice-president and enthusiastic evangelist, has been on a road trip, meeting with tech writers and analysts to persuade them of the evils of MobileMark.
As a group, I don't think we really need much persuading. I look at lots of laptops and find that their battery life claims bare rather less relationship to reality than the environmental Protection Agency's mileage ratings do to what you can actually expect from a car. MobileMark, the product of an industry alliance, test notebooks under thoroughly unrealistic conditions, including turning screen brightness down to 20% of maximum.
I wish I could give readers better information on battery life. Mt friend and competitor Walt Mossberg runs what he calls a "harsh battery test" in which he times how long a laptop will run with the brightness turned all the way off, all power savings shut off, and playing an endless loop of music. But I think this sets up conditions not much more realistic than MobileMark and, by turning off power saving, fails to give laptop designers credit for some impressive work they have done to manage power consumption.
I like Moorhead's suggestion that the laptop industry follow the example set by Apple, which rates the battery life of the iPhone for a number of different conditions such as talk time, Wi-Fi Web browsing, and music playback. A benchmark that said you could expect 4 hours of battery life working in Microsoft Word or Outlook, 3 1/2 hours actively browsing the Web, and 2 1/2 hours playing a DVD would be truly useful. How about it, BAPCo?