Hacker News Reader

Comment by lxgr | Hacker News Reader

Renewables don't exist in a vacuum: "Green" miners buying up capacity drives up the price, driving others into non-renewables as a consequence (if only temporarily).

Until we are at 100% renewables, supporting a more energy-consuming product/solution/process while alternatives readily exist should not be called "green", in my opinion.

View on HN

Bitcoin mining operations are generally located where electricity is cheapest. Electricity is cheapest where there is a surplus of hydro power (or other power with very low marginal cost). Where there is a surplus, use of that surplus does not drive other customers to other power sources.

By that logic, Bitcoin disincentivizes any investment in better grids and more efficient energy storage.

Also, any such mining rig could have reasonably been a data center or scientific supercomputer instead.

That's not "logic", that's reality. What you said is pure speculation.

Indeed, speculation offering a counterpoint to the implicit assumption that renewable energy used by Bitcoin miners would not have any capability of impacting the energy market when used more efficiently.

I think your second point here often gets missed/understated. That energy and those computers could have been helping forecast severe weather events or working to fold proteins or any other number of useful tasks that don't involve tracking a ledger of other people's wealth.

The energy perhaps. The "computers" aren't. They're just ASICs that can calculate hashes at a phenomenal rate, they're not general purpose.

And the argument about energy isn't straightforward either. Most bitcoin mining sites are set up where there is an excess of energy - when I had miners a while ago, I used a place in Newfoundland, Canada to host them, because of the local hydro dam. The water falls through the sluices or goes via the generators, it's just based on demand.

>The energy perhaps. The "computers" aren't. They're just ASICs that can calculate hashes at a phenomenal rate, they're not general purpose.

The chips are made with recent-ish nodes, so the silicon/fab capacity could theoretically be used to make CPUs.

edit: looks like at least one bitcoin ASIC maker is using TSMC 7nm, the same process used to make AMD CPUs.

[1] https://blog.bitmain.com/en/bitmain-announces-next-generatio...

>could have reasonably been a data center or scientific supercomputer instead.

Which would have been used for what? Supercomputer centers are having difficulty selling their cpu cycles. A data center for what? Figuring out how to better serve ads?

Before you say protein folding, I was in on meetings where s major pharma company was considering decommissioning their computational data center in the bay area because it was underused and the real estate was worth too much to leave on the table

Doesn't that take away from useful aluminum production?

> Renewables don't exist in a vacuum

Yes they do, like all energy sources there's a physical limit on how far you can transmit it, all power generators live in a regional bubble.

There's a reason all that excess hydro energy in Sichaun province goes into negative prices multiple times a year and can't be bought elsewhere in the country or even exported to other countries, it's simply too far away.

> it's simply too far away.

Brazil has a 2500 km long ultra high voltage direct current line running from a hydroelectric power plant to Rio de Janeiro. I'm pretty sure there's a medium-sized city or two within that radius of Sichuan.

That's the longest in the world, not exactly representative of the norm. There's grid bubbles everywhere, they do all live in a "vacuum", you can't just magically transport it without decades long mega infrastructure projects to build super grids and accepting large transmission losses.

Reply on news.ycombinator.com