Sia hard fork bricks Innosilicon and Bitmain ASIC miners as punishment
Siacoin is hard forking to ruin competing miners, while leaving its own Obelisk miners fully functional.
An equally exciting and troubling development has gone down today, with Siacoin passing through a hard fork that destroys competing application specific integrated circuit (ASIC) mining machines, turning them into so many tens of millions of dollars' worth of useless paper weights. Specifically, the fork will be bricking Innosilicon and Bitmain Siacoin miners.
At the same time, Obelisk miners are left unaffected. The same people run both Siacoin and Obelisk, so this decision is essentially just Siacoin handing its own partner company a complete, highly centralised monopoly over the Siacoin network.
It's well worth unpacking this fascinating development, and its implications, for quite a few reasons.
Piece by piece
Proof of work is highly centralised by nature
There were several different motivations behind the fork, and it wasn't one that the Siacoin team took lightly. In fact, the issue first came up months ago but was decided against largely because people were worried about what would happen if Siacoin's creators got a mining monopoly over their own network.
But as David Vorick, lead developer of Sia and CEO of Obelisk said, almost any proof-of-work cryptocurrency will inevitably find itself under a mining monopoly. The alternative is GPU minability and anti-ASIC forks, but these are unreliable for two main reasons.
The first reason is because it will only ever be a stopgap measure and means committing to an endless series of anti-ASIC forks. The second is because it makes a coin a ripe target for 51% attacks through rented mining alone or through secret ASICs.
"To the best of my knowledge, no ASIC mined coin has ever been hit by a 51% attack. And yet, something like a dozen different GPU mined coins have been hit by 51% attacks, typically to steal from exchanges," he said in the official announcement.
ASIC mining, meanwhile, means accepting that all but the largest cryptocurrencies – bitcoin and that's about it – will become highly centralised under a monopoly. This is down to the market forces of ASIC mining, he says.
It requires a large upfront investment of at least $5 million to start manufacturing ASIC miners for a certain algorithm, which means manufacturers need to be confident that there's enough money waiting for them. Unless you're making a competitive mining machine for bitcoin or another very large cap coin, this means finding a market you can dominate by rendering the competition obsolete.
Because mining companies aim to recoup their investments as soon as possible, and know that their own hardware will eventually be usurped by newer machines, the smart thing is to milk a coin's community for as much money as possible as quickly as possible. The goal is to monopolise a market and then use that monopoly to gouge as hard as possible. If people don't want to pay the price for the machines, a mining company can always just use them itself.
"For a new product that is going to obsolete all existing hardware and become the new monopoly, the market size required to justify investment is in the mid tens of millions of dollars. But for a new product that is going to merely break up an existing monopoly rather than replace it, the market size required to justify investment is hundreds of millions of dollars due to the much reduced margins and presence of competition," Vorick explains.
"Based on what we’ve seen out in the market, the total amount of hardware sales that a cryptocurrency can support is roughly equal to its annual block reward. Cryptocurrencies with a block reward below hundreds of millions of dollars per year will not be able to profitably sustain multiple ASIC manufacturers. This cleanly explains why Bitcoin is the only cryptocurrency with multiple reasonably competitive machines."
Mining as a service
Vorick sees miners as service providers for cryptocurrencies. They provide the service of protecting the network and are in turn paid millions of dollars per year for providing this service. Communities have every right to hold ASIC manufacturers to a high standard, he says.
"All told, we believe that embracing an ASIC monopoly is better than resisting ASICs altogether," Vorick says. "That said, we don’t believe that a cryptocurrency needs to embrace a parasitic or abusive ASIC monopoly. ASIC manufacturers ultimately exist to serve the network, and specifically to protect the network against 51% attacks. Even small cryptocurrencies pay millions of dollars per year for this protection, and it's reasonable to hold any incumbent manufacturer to high standards."
It's obviously a little naive though. As Vorick said, market forces favour a strategy of monopolisation and gouging, and a mining company that adopts a more altruistic paradigm probably won't last long.
Enter the kill switch
Vorick sees the solution to parasitic miners as built-in kill switches, which allow for the selective destruction of competing machines. In the case of Obelisk's Siacoin miners, the kill switch was presciently built in right from the start, as "a tiny, secret extra circuit" which "would allow the Sia developers to perform a hard fork that breaks manufacturers without that extra circuit, but does not break the Obelisk machines... we encourage all manufacturers to add extra secret circuits that can act as failsafes."
Revenge served hot
This is not a nice hard fork. It's going to destroy tens of millions of dollars of hardware and make a lot of people very unhappy. And on one level, it's a straightforward act of revenge. Obelisk was in the process of publicly making its own Siacoin miners, and had already invested millions to that end, when competitors came out with machines that rendered Obelisk miners obsolete.
"An important line was crossed when secret ASIC projects superseded a public project that had substantial community investment," Vorick writes. "Sia is forking today to reprimand the current ASIC monopoly for the damage it did to the Sia community, to make whole the supporters of Sia's community ASIC project, and to send a clear message to all future Sia ASIC manufacturers: we will not tolerate an abusive ASIC monopoly.
"In v1.3.6, Sia will be releasing hardfork code which bricks the Bitmain and Innosilicon ASICs, but continues to allow Obelisk ASICs to mine on the Sia network. This revokes the Innosilicon monopoly and instates Obelisk as the incumbent ASIC monopoly."
This is a radical departure from previous decentralisation paradigms. It's a cryptocurrency whose developers manufacture their own ASIC miners with kill switches that let them destroy rivals. Siacoin is unique in this respect.
Vorick sees the need for a hard fork as a kind of failsafe inside that, whereby no forks can take place without the approval of the community. Of course, when a proof-of-work coin talks about community approval, they're really talking about miner approval – they're the ones who decide whether the upgrade goes forwards.
A developer might project fluffy altruistic ideals onto their communities, but Obelisk miners aren't any more altruistic than Bitmain or Innosilicon miners. They're in it for the money just as much as anyone else, and this hard fork is a protectionist move to ensure their own return on investment at the expense of their rivals.
If all manufacturers start building in their own kill switches – and many undoubtedly already are, they're just being quieter about it than Obelisk is – you might get an entirely new proof-of-work system where monopolies are transparently passed on through these kinds of destructive forks.
This won't do anything to prevent abusive monopolies, and Vorick also notes that you can't just keep bricking miners without compromising network security, but at least some additional transparency might mean people will stop pretending that proof of work in any coin, except perhaps bitcoin, is decentralised.
Plus, by virtue of the sheer hardware expenses and energy consumption involved, ASIC proof-of-work mining will always be purely about the money.
The most likely outcome for Siacoin in the long run is that Obelisk as a company changes hands or goes public, and then starts using its monopoly to gouge the Sia community. You can't eliminate basic greed with a hard fork.
That's assuming proof of work lasts long enough for Obelisk to go bad. As this entire fiasco shows, it's a wasteful, deeply flawed and inherently highly centralised type of consensus mechanism, which is completely unsuitable for serious use in the long run. Alternatives might not be ready for prime time yet, but a lot of people will probably be glad to get off the PoW ride when they are.
Disclosure: At the time of writing, the author holds ETH, IOTA, ICX, VET, XLM, BTC and ADA.