# The weight of the Internet

Credits: Charles Arthur, Technology section of ‘The Guardian‘.

The Internet, what it is now, weighs either about $60\, g$, or about $6\, \mu g$ according to the method you use to weigh it.

The method takes a guess at the number of servers running the net (between 75m and 100m), their average power consumption (between 350W and 550W), the average voltage inside a logic gate (3V), and the average speed of those chips (1GHz).”An ampere is some 1018 electrons a second,” Seitz writes, totting up the power use at 40bn Watts (40GW): “Straightforward calculation reveals that some 50 grams of electrons in motion make up the internet.” Always bearing in mind that each electron has a resting mass of $9.1 . 10^{-31} kg$, of course – so it takes a lot of them to make up even that tiny weight.

Above method is due to Russell Seitz.

Another method that predicts the weight to be a mere $6\, \mu g$ is due to Stephen Cass:

He used the weight of a “bit” – comprised of 40,000 electrons stored in a capacitor on a chip. Bear in mind that the average 8-bit byte only contains four “1” bits (and four “0” bits), multiply it by the total volume of information passing around the net, estimated at 40 petabytes, and voila: 0.2 millionths of an ounce. Or so. Of course, once your electron starts moving, its weight will rise (due to relativistic effects). So perhaps the net really does get slower as more people use it.

That was published by the Discover magazine.

Advertisements

## 2 thoughts on “The weight of the Internet”

1. linpaws says:

Tell me the use of this information…sounds interesting…may be if we’re running fusion reactors, we might talk about saving BW to help conserve on weight so that it could be transformed to energy! Is it a direct hit on energy consumption by the internet BW usage?

2. I believe the calculations done in both the cited articles are erroneous. Both of them do not take into account the GPRS and 3G data. But yes, compared to the magnitude of the “traditional” Internet data, that is negligible for the first calculation; for the second one (6 microG), it could be substantial, for the way the calculations are done in it.

I’m not sure how electrons help in conserving energy. One use I can think of right now is in devising plans for an atomic warfare through Internet cable. If somehow we could use electrons in “fission”, “fusion” or some other sub-atomic non-chemical reaction with silicon or GaAs involved, we can conquer the world.