Fun thing about evolution. Suppose you consider every cell division a calculation for evolution (which it is). A good global average for cells is ~2 hours betw

Fun thing about evolution. Suppose you consider every cell division a calculation for evolution (which it is). A good global average for cells is ~2 hours between cell divisions (because plants, animals, ... are utterly insignificantly small parts of the biosphere, they don't matter at all). How many cells ? Well one can assume that every spot on earth, on average has ~2 tons per cubic meter of continuous biomass on it. In the oceans this is very uniformly true (and more like 3-4 meters but it decays exponentially as you go down), and on land the topsoil is much the same, but not as deep. A back of the envelope figure for biodensity is 1.210^8 cells/g. So there would be ~2 tons of biomass per square meter of earth, making 2.410^14 cells per square meter, for 510 million km2. That makes something like 10^30 cells total for the earth, each of which computes on average once every 2h. Or 1.4e26 computations per second (and these are full "does this work ?" simulations, not a single addition). > A paper published in Science estimated that in 2007 the total global computing power was 6.4×1018 instructions per second. This means that evolution tries a billion new survival strategies for every assembly instruction any computer worldwide executes.