3 ms·

I think Chia makes a pretty good case here — same Nakamoto consensus security guarantees as Bitcoin only using hard drive space instead of processor power so th

by Glench 5y ago

I think Chia makes a pretty good case here — same Nakamoto consensus security guarantees as Bitcoin only using hard drive space instead of processor power so the ongoing energy use is vastly less. There are around ~350k full nodes today (compared to bitcoin's 15k) and it uses about 400x less power: https://chiapower.org/ https://chiapower.org/ I farm it completely with old hard drives that I had lying around that weren't being used anyway.

cuteboy19 5y ago
So instead of wrecking the GPU market we would move to wrecking the HDD market?
- Glench 5y ago
  I'm not sure if you're interested in discussing or just want to do a drive-by criticism, but that's actually an interesting question! Currently 1.5 zettabytes of new HDDs are sold each year which is around $65 billion. So HDDs are a pretty mature market with massive data centers buying most of that storage. One of the hypotheses with Chia was that there is a lot of over-provisioned storage already out in the world — on both consumer and idle enterprise drives. This is in line with the original intent of Bitcoin that anyone could mine which would make the network more secure. It's obviously cheaper to use existing storage to secure the Chia blockchain than to buy new drives so the market will move toward maximally using existing storage. Datacenters could use their existing idle storage for Chia farming before a more lucrative use came along. And the storage doesn't need to be very reliable for Chia farming, so old hard drives that are no longer viable from datacenters can be used (apparently datacenters just destroy HDDs outright currently?) Currently there are about 35 exabytes of storage in the Chia network, which is about 2-3% of the entire yearly supply of HDDs. Given that supply chains weren't really disrupted, I think it's pretty safe to say that the Chia bet was right and that it's actually utilizing a bunch of unused space already out in the world. And the IDC agrees: https://idcdocserv.com/US47836321 https://idcdocserv.com/US47836321 Now, obviously if the market valued Chia extremely highly then supplies of HDDs would be impacted. But I wonder how much that would affect everyday folks like you and I since most storage in devices are now SSDs, which are not very storage-dense so aren't really as economical for Chia. Would it be the case that consumers couldn't buy HDDs? At some coin price, yeah, definitely. I have no idea what that price could be.
tromp 5y ago
Chia has weaker guarantees than Nakamoto consensus. Every proof-of-work solution in Bitcoin commits to the transaction history, but the proofs-of-space in Chia only commit to the farmer public key. Which is used to separately sign for a block of transactions. Which allows a collusion of recent farmers to produce alternate transaction histories at no cost.
- Glench 5y ago
  You could be right! Do you have any evidence, research, or other supporting arguments that show what you're claiming?
  - tromp 5y ago
    See https://docs.chia.net/docs/03consensus/foliage/ https://docs.chia.net/docs/03consensus/foliage/ "farmers could theoretically create a foliage re-org where foliage is replaced" As you can see, Chia devs consider this an impractical attack. But it's definitely a weaker security guarantee than what PoW provides.
    - Glench 5y ago
      Ah, thanks for the link! It seems what you're saying is right — the security guarantees of Chia are not the EXACT same as Bitcoin, but it's not clear to me that this means they're "weaker" for the reason you described. > As soon as one honest farmer has added a foliage block, the foliage becomes impossible to re-org beyond that height with the same PoSpace With ~350k Chia full nodes, the likelihood of colluding farmers being able to pull this off is a lot lower than if there were only the ~14k full nodes of Bitcoin, no? And Chia is still pretty new — it had a maximum of ~750k full nodes before. I guess it's kind of a trade-off, huh? Bitcoin's big centralized pools are a worrying source of attack that's not there in Chia (the official pooling protocol has the farmers sign blocks so it's completely decentralized and pools using it can't orchestrate attacks). https://docs.chia.net/docs/11pooling/pooling https://docs.chia.net/docs/11pooling/pooling
jquery 5y ago
Huh. I have nearly 100tb in unused but brand new hard drives. Might be worth using them to help create an energy efficient alternative to bitcoin.

by Glench 5y ago

cuteboy19 5y ago

So instead of wrecking the GPU market we would move to wrecking the HDD market?

Glench 5y ago
I'm not sure if you're interested in discussing or just want to do a drive-by criticism, but that's actually an interesting question! Currently 1.5 zettabytes of new HDDs are sold each year which is around $65 billion. So HDDs are a pretty mature market with massive data centers buying most of that storage. One of the hypotheses with Chia was that there is a lot of over-provisioned storage already out in the world — on both consumer and idle enterprise drives. This is in line with the original intent of Bitcoin that anyone could mine which would make the network more secure. It's obviously cheaper to use existing storage to secure the Chia blockchain than to buy new drives so the market will move toward maximally using existing storage. Datacenters could use their existing idle storage for Chia farming before a more lucrative use came along. And the storage doesn't need to be very reliable for Chia farming, so old hard drives that are no longer viable from datacenters can be used (apparently datacenters just destroy HDDs outright currently?) Currently there are about 35 exabytes of storage in the Chia network, which is about 2-3% of the entire yearly supply of HDDs. Given that supply chains weren't really disrupted, I think it's pretty safe to say that the Chia bet was right and that it's actually utilizing a bunch of unused space already out in the world. And the IDC agrees: https://idcdocserv.com/US47836321 https://idcdocserv.com/US47836321 Now, obviously if the market valued Chia extremely highly then supplies of HDDs would be impacted. But I wonder how much that would affect everyday folks like you and I since most storage in devices are now SSDs, which are not very storage-dense so aren't really as economical for Chia. Would it be the case that consumers couldn't buy HDDs? At some coin price, yeah, definitely. I have no idea what that price could be.

tromp 5y ago

Chia has weaker guarantees than Nakamoto consensus. Every proof-of-work solution in Bitcoin commits to the transaction history, but the proofs-of-space in Chia only commit to the farmer public key. Which is used to separately sign for a block of transactions. Which allows a collusion of recent farmers to produce alternate transaction histories at no cost.

Glench 5y ago
You could be right! Do you have any evidence, research, or other supporting arguments that show what you're claiming?
- tromp 5y ago
  See https://docs.chia.net/docs/03consensus/foliage/ https://docs.chia.net/docs/03consensus/foliage/ "farmers could theoretically create a foliage re-org where foliage is replaced" As you can see, Chia devs consider this an impractical attack. But it's definitely a weaker security guarantee than what PoW provides.
  - Glench 5y ago
    Ah, thanks for the link! It seems what you're saying is right — the security guarantees of Chia are not the EXACT same as Bitcoin, but it's not clear to me that this means they're "weaker" for the reason you described. > As soon as one honest farmer has added a foliage block, the foliage becomes impossible to re-org beyond that height with the same PoSpace With ~350k Chia full nodes, the likelihood of colluding farmers being able to pull this off is a lot lower than if there were only the ~14k full nodes of Bitcoin, no? And Chia is still pretty new — it had a maximum of ~750k full nodes before. I guess it's kind of a trade-off, huh? Bitcoin's big centralized pools are a worrying source of attack that's not there in Chia (the official pooling protocol has the farmers sign blocks so it's completely decentralized and pools using it can't orchestrate attacks). https://docs.chia.net/docs/11pooling/pooling https://docs.chia.net/docs/11pooling/pooling

jquery 5y ago

Huh. I have nearly 100tb in unused but brand new hard drives. Might be worth using them to help create an energy efficient alternative to bitcoin.