Even 'Free' Chia Farming Wouldn't Be Profitable at Scale

tomshardware.com

21 points·GordonS··7 comments

It seems like this is bad for new farmers and some existing farmers will eventually give up, but it should reach an equilibrium eventually?

Also, it's good that it won't put much demand on SSD prices at that point.

Yeah the article doesn’t make sense. It seems to assume exponential demand increases without the same increases in pay.

Crypto-mining of any kind doesn’t work that way.

Either the article/backblaze is wrong or Chia as a whole is going to quickly collapse.

> To quickly understand what costs would look like, we used our standard pricing of $5/TB/month as our effective 'cost'

Price is not cost! Far from it (I hope).

Not sure why backblaze used ssd numbers when you can run cheaper magnetic drives:

https://www.reddit.com/r/chia/comments/odox2x/working_on_my_...

iirc, because Chia mining only works with SSD and better

There is a bit more to it. Chia "plots" are large files that act as bingo cards. You need an enterprise-grade SSD to generate a card but not to play bingo once you have it.

The files being expensive to create but not to reuse is what makes it proof of space, since it makes caching the data much cheaper than generating a new card.

Semantics aside, you require a SSD or better at some point and that was where Backblaze was seeing above average failures and started reining in Chia usage iirc

SSDs (or NVME) are used for plotting generation - but serving them can be done with big, dumb, slow spining drives.

theres still no backing to any crypto that stabilizes its value, so you can basically make up any valuation you wish.

Reply on news.ycombinator.com