SoylentNews is people
SoylentNews
According to a press release carried by Eurekalert
In the first rigorously peer-reviewed article quantifying Bitcoin's energy requirements, a Commentary appearing May 16 in the journal Joule, financial economist and blockchain specialist Alex de Vries uses a new methodology to pinpoint where Bitcoin's electric energy consumption is headed and how soon it might get there.
The abstract of the article says
The Bitcoin network can be estimated to consume at least 2.55 gigawatts of electricity currently, and potentially 7.67 gigawatts in the future, making it comparable with countries such as Ireland (3.1 gigawatts) and Austria (8.2 gigawatts). [...]
The author offers a caveat:
[...] all of the methods discussed assume rational agents. There may be various reasons for an agent to mine even when this isn't profitable, and in some cases costs may not play a role at all when machines and/or electricity are stolen or abused.
[Other] reasons for an agent to mine Bitcoin at a loss might include [...] being able to obtain Bitcoin completely anonymously, libertarian ideology [...] or speculative reasons.
(Score: 2) by JoeMerchant on Monday May 21 2018, @07:28PM (2 children)
You missed an order of magnitude, or four... a single transaction in bitcoin currently consumes the equivalent of $100 in electricity in my neighborhood. If my world ran on bitcoin, a $200 load of groceries would be taking $100 in transaction cost overhead (assuming that the magic security takes care of all the current processing fees, which it won't) as compared the the current $2 that my credit card processor skims out of the transaction.
🌻🌻 [google.com]
(Score: 2) by VLM on Monday May 21 2018, @08:46PM (1 child)
I would suspect a booming business in the future for fractional reserve bitcoin transactions. And small local altcoins.
I would not be surprised if the future is a patchwork of altcoins and automated middlemen and the altcoins are some variation on fractional reserve or full reserve banking.
The entire world MIGHT need to sign that there's a major international trade agreement, but expecting the whole world to sign my bag of lettuce yesterday afternoon is unlikely in the future. On the other hand my local regional 16 store supermarket chain might have a coin signed at every store and the "real" bitcoin market is only contacted occasionally to keep in sync. That might be too small. My credit union belongs to a union of credit unions such that I get free ATM use at a ridiculous number of ATMs, and that union of credit unions could have an alt-coin of their own, to run maybe 25% of the economy of one major metropolitan area. Big (fee) banks like wells fargo have thousands of retail sites, I could see them having an alt-coin all their own.
If you want the entire world to sign your financial transaction, its gonna cost, but merely having 100s of metro area credit unions sign is actually quite cheap.
Something not often discussed is watt/MIP and $/MIP have been improving for a long time plus or minus Moore's Law ending soon, etc. So much as my first desktop computer in '81 had 16K of memory compared to my 32G desktop today, possibly the cost of bitcoin in energy terms will drop in 40 years to "nothing". Unlikely but possible.
Another interesting way to measure the power consumption is looking at solar power; if the planet were covered by a mesh of solar powered bitcoin nodes, its not really a problem as long as enough are in sunlight at any given time. Nobody really cares how much electricity it would take to generate light using LED lightbulbs to grow all the worlds corn, although I'm sure it would be a huge power draw if it were tried. Often when discussing solar powered computation you get complaints its impossible 24x7 blah blah blah, well clustering software simply transfers load where there's sunlight, and zero power bill means the software dev cost is irrelevantly low.
(Score: 2) by JoeMerchant on Monday May 21 2018, @09:34PM
Proof of work is some self fulfilling prophesy as far as cost vs value... Whatever mining "costs" is about what miners will be willing to mine (or transactors are willing to transact, if you will) for the built-in reward of doing the operation. I read somewhere (terribly reliable, I'm sure) that BTC miners "produced" $8B in BTC over some 12 month period, and spent ~$3.5B in electrical power costs to do so. This entirely neglects the cost of mining hardware, housing for said mining hardware, labor to install and maintain, internet connection costs, etc. So... the "free" market demonstrates its power of balance at scale yet again: as long as (speculative idiot) investors are paying $10K/BTC, mining will ramp up to grab that cash - nearly to the breakeven point. And, due to the scaling difficulty feature of the BTC algorithm, there's no end to how much power can be brought online, as long as "investors" are paying for it.
If we manage to scale back the cost per transaction down to the 1-2% of transaction value range, what have we won? The ability to tell Visa, MasterCard and Amex to F-off? Not really, not for a long time. Sure, you can create blockchain systems that peg the value of a "coin" to a dollar, or euro, or whatever - but that brings the big central figures back into the picture, not to mention regulation and overhead to ensure that the backing funds are appropriately administered, etc. Until that semblance of stability is present, no 16 store supermarket chain is going to accept a "currency" with wild hour-to-hour fluctuations in value vs the products they buy and sell.
I agree, there is a future where widely held cryptographic trust systems replace a lot of what "money" does today, but I do not agree that proof of work (mass hashing of nonces to regulate publication speed) is part of that future. Even if watt/MIP and $/MIP fall through the floor, that will just drive the POW complexity requirements upwards - it's peoples' behavior that's driving the cost of proof of work, not the cost-complexity of the work itself.
🌻🌻 [google.com]