Hacker News Reader

Comment by paulryanrogers | Hacker News Reader

Can anyone speak to the possible energy savings this change may provide?

Or is BTC too fundamentally tied to CPU-bound work?

There will be energy savings for the nodes that have a copy of the blockchain and need to verify transactions but not for the miners.

Market forces fundamentally tie the energy consumption of the miners to the price of bitcoin.

You can't save energy in bitcoin, because proof-of-waste is intrinsic to it's functioning and the difficulty adjusts to keep it that way.

Bitcoin miners are utilizing excess energy that would otherwise have gone to waste. So there's nothing wasteful about it. In fact, miners must be as energy efficient as possible in order to remain competitive and profitable. Typically excess power is the cheapest source because there's no demand for it, hence why efficient miners use it. Check out great American mining[1] for example. They harvest wasted energy from gas flares to power bitcoin mining rigs.

1. https://gam.ai/

Even with free electricity, the embodied energy of the computers used is basically wasted. A simple notary service that signed transactions with a timestamp and published a record to a public data storage bucket could handle worldwide bitcoin transaction volume running on a single machine, if you traded proof-of-waste for a simple root trust store.

(I mean, if I wrote it it would probably take several hundred machines but I'm sure Satoshi or some other brilliant programmer could fit it on a single machine. It only processes what, 5 transactions per second? My first dumb cellphone might be able to do that many signatures per second if someone could figure out how to compile the relevant libraries to run on it.)

Goodluck with that idea. I'll be sure to check back on this next year to see if anyone has made progress on it.

Personally, I prefer to use a variety of proven private ledgers with a variety of trust roots, independent audits, and insurance. Using a public ledger is not something I would invest in. I'm just saying Bitcoin is a wildly, incredibly inefficient public ledger service.

This is basically false (some miners might do this but they're a minority). Almost all bitcoin mining happens in China, and happens in an area where a large fraction of grid power comes from coal. Using excess power doesn't work well for bitcoin mining because the power draw is constant: miners are generally not powered on or off over the coarse of a day: they're turned on and not turned off until they're decommissioned.

Bitcoin miners are the ones that use the energy calculating the SHA256 functional. The cost of the function is not relevant to energy use as competition between miners means the energy used will match the value of the block reward.

This optimizing is for the users of the bitcoin network that must validate that the rules are being followed so they can reject any miners that do not follow the consensus rules

This only helps nodes verify blocks, the mining is what uses huge amounts of power and that won't change.

ecdsa speed is related to client performance, it has nothing to do with the power consumption of proof of work.

This is not relevant to bitcoin mining. Just synchronization.

The mining is no longer (for 7 years or so) done by traditional CPUs. People use ASICs

Note that this is not mining, just running a full node. When running one, signature verification is most of what your CPU is doing.

Reply on news.ycombinator.com