Do the numbers in that article, uhhh, add up? The article estimates the total number of ATMs at 3 million, with an average daily power consumption of 5Wh, for

Do the numbers in that article, uhhh, add up? The article estimates the total number of ATMs at 3 million, with an average daily power consumption of 5Wh, for a total yearly power consumption of 13Twh. Except if I punch [3 million * 5Wh * 365.25 to terawatt-hours] into the Google, I get 0.0055TWh. If I assume the author meant to type 5kWh as the average daily power consumption, you get 5.5TWh, which is closer, but still not the stated quantity. [EDIT: OK, so I if I punch [3 million * 500W * 365.25 days to terawatt-hours] into the Google I get the 13TWh the author estimated for ATM consumption. So the author is actually assuming that the average ATM is consuming 500W?] Likewise the numbers for how much electricity bank branches consume doesn't seem to add up (it seems like the author assumes branches consume an average of 10kW, not 10kWh/day?), and I can't replicate any of the numbers given for bank server energy consumption (just multiplying the numbers given in the article together in different ways). Can someone else get the numbers in the original article [https://hackernoon.com/the-bitcoin-vs-visa-electricity-consumption-fallacy-8cf194987a50 https://hackernoon.com/the-bitcoin-vs-visa-electricity-consu...] to add up?