mehack

Raffi, i enjoyed your talk a lot!
Your calculations only seem to take into account, what is emitted by using the twitter infrastructure.
Not taking into account the CO₂ emitted through the use of the computer used to compose the tweet, the internet as a transport medium to distribute the tweet, 3rd party servers like twitpic or amazon S3.

So taking all this into account, what do you believe a good, overall number would be per tweet?

Google claims a search emits 0.2grams of CO2 which is already tenfold the number you came up with.

I'd love to hear a good over all number, so i can adjust the calculations on http://stefanseiz.com/twitter-co2.html

Cheers,
Stefan

you have to draw the line somewhere when doing these sorts of energy consumption calculations.  this number reflects pumping a tweet through the infrastructure, and yes, does not account for storage on S3, internet infrastructure (or cellular infrastructure if you are coming from mobile), the device that you are tweeting from, etc.  in reality, i would like to extend this out to also account for the travel that Twitter employees do, the power it takes to power our offices, the laptops we have at the office, etc.  

Raffi - great job showing others how they need to think about energy per unit of IT work. Here is my own blog post on Green Data Center blog http://www.greenm3.com/2010/04/carbon-footprint-of-a-tweet-energytweet-approx...