This PR reintroduces the constants to use base 10 figures for calculations, so when we say Gigabyte we are referring to 1000 * 1000 * 1000 bytes.
This is assuming the the original paper, when it refers to transfer in terms of gigabytes, is using the same calculation.
What should this do to the calcs?
This should have the effect of making emissions figures for every calculation go up slightly, as the top down model that this is based on divides a set amount of energy use allocated to an entire system, then divides it by a very large number of gigabytes of transfer, to figure out a energy used for give amount of transfer.
By making the number we divide the total energy by smaller, we end up with higher energy use per byte, so we end up with higher numbers for carbon emissions as well.
This PR reintroduces the constants to use base 10 figures for calculations, so when we say Gigabyte we are referring to
1000 * 1000 * 1000 bytes.This is assuming the the original paper, when it refers to transfer in terms of gigabytes, is using the same calculation.
What should this do to the calcs?
This should have the effect of making emissions figures for every calculation go up slightly, as the top down model that this is based on divides a set amount of energy use allocated to an entire system, then divides it by a very large number of gigabytes of transfer, to figure out a energy used for give amount of transfer.
By making the number we divide the total energy by smaller, we end up with higher energy use per byte, so we end up with higher numbers for carbon emissions as well.