Does anyone know the numbers and models by which Charles estimates the number of 5W RockPIs it would take to run the whole network of Cardano nodes. I’m guessing that will go up with transaction rate. Would that be at current transaction rates? Rates at full launch? Even if this doesn’t have an exact answer, a back of the envelope sketch.
It makes a fun network metric to compare with other chains. BTC would fail here for sure.
Yes a busy machine consumes more power, but power ratings generally indicate the maximum consumption, and these are ballpark figures anyway, so I’d say just multiply 5W by the number of nodes.
Yes, that part is trivial. I mean about how many nodes you would need. If he means in the current configuration, and if so, how do we expect it to scale with traffic. Just saying that someone probably has some numbers and models we could use as an upper limit until actual experience develops. Whether it consumes exactly 5W running the node isn’t important, but whether it can’t really cope …
In any case, presumably there would be a technical argument for running very powerful nodes at the core of a fully scaled network. I’m a systems engineer, so I want to see this part of the theory and development models just as much as the security and other analysis. I can see where having a full node on $100 hardware at the edges of the network could serve a lot of low/no tech users as well. Old smart phones are about the same power computationally, which is the more important deployment case.