Flash Websites Use More Electricity Than Non-Flash Sites?

Filed under Industry by Chad | 3 comments


A post over at O’Reilly leads to a blog post and a *ahem* white paper on a topic that seems a bit self-explanatory, albeit sensationalistic. Obviously, this paper is meant to stir the pot a bit. There are a number of ways he could even out his research methodology to make it seem less like an AJAX developer/GLU (Grumpy Linux User) who just doesn’t like rich media and more like a researcher that writes white papers… Some key points:

  • MySpace? You are basing all of this on one domain? One site where individual profiles are notoriously messy and strewn with nastiness? What about Youtube, Facebook, Cnet, Digg, etc. A broader sampling should probably be used to get an accurate view.
  • How about comparing Flash to Silverlight, Java or heavily AJAX-using sites?
  • One version of Flash player? Who knows if Flash player 9.0.115 is more or less efficient than previous revisions.
  • One browser? Hmm.
  • One OS? What was it, Linux? Again, hmm.
  • No mention of other services etc running on the machine.
  • No mention of power management settings in his OS control panels.

Now, I’m all but sure that Flash enabled sites do indeed use more juice than non-Flash sites… It just makes sense, as my fan rarely kicks on when surfing, except when visiting Flash sites like YouTube, etc., however, if you are going to set up your findings to appear as a scientific experiment, it really should be a bit more thorough. I realize that I sound a bit like a wet blanket here… but ringing the alarm bells seems a bit premature.

I would welcome reading a complete, detailed report containing a broader sampling with a more comprehensive methodology. That said, I applaud Matt for taking some steps to try and research Adobe’s black box Flash player and since it appears he is certainly really smart, and looking for a job, perhaps consulting as an energy efficiency engineer for Adobe’s Flash player team would be a good gig for him. ;-)

[del.icio.us] [Digg] [dzone] [Facebook] [Google] [Newsvine] [Reddit] [StumbleUpon] [Technorati] [Twitter] [Yahoo!]

Posted on January 24, 2008


3 comments

  1. Arnaud Jan 24

    On the other hand, client technologies like Flash can reduce the bandwith compared to a traditional HTML site because we don’t have to refresh all the page, so all depends on the type of application we compare…

  2. John Dowdell Jan 24

    Your points on methodology are well-founded, at least to support the headline made at O’Reilly.

    But I found the original PDF amusing… academic tone, background digressions up top, you have to analyze the piece hard to see what the author is trying to say and what leads them to say it, and whether it’s believable. (Thanks for itemizing the missing requirements, Chad! ;-) Seems like the original question was more “What’s the environmental impact of all these heavy amateur sites?” than what the subsequent attack portrayed.

    It would be a good topic to pursue — with the world embracing web video, and the move from telephony to internet telephony, then what marginal costs are we accruing? How do they correspond to, say, escalators or overheated buses? It’s a valid question to ask.

    jd/adobe

  3. Matthew Garrett Feb 5

    Right – the intention wasn’t to bash Flash, it was to work out what the extra cost of letting people embed whatever they want in websites. It turns out that it’s noticable but not huge. If people actually want to save the world then there are massively more significant things they could do, but I think it’s interesting that personal website design has an effect on other people’s electricity bills.

Leave a reply

Related Posts:

  • links for 2009-04-21
  • About Me
  • Adobe ripped off a blogger to create Kuler?