A few days ago, I was chatting with a friend who has dozens of 3090 graphics cards that were previously used for mining Ethereum and have been idle since the merge. I told him, you can try the Boundless prover network, and he looked completely confused.
The reasoning is quite simple. After Ethereum transitioned to POS, a large number of GPU mining rigs were left without a purpose. Although some have switched to mining other coins, the returns have been quite mediocre. AI training does require GPUs, but the entry barrier is too high for regular miners to participate. The proof generation of Boundless perfectly fills this gap.
From a technical perspective, generating ZK proofs is somewhat similar to mining, as both involve a lot of parallel computations. However, the difference is that the hashes calculated during mining are useless, while each generated proof has actual value. It's like turning wasted electricity into useful computing resources.
The design of Boundless's Proof of Verifiable Work mechanism is very clever. It's not about who has more computing power, but rather the complexity of the proof you generate. A complex proof might earn more than ten simple proofs. This encourages provers to optimize algorithms and improve efficiency rather than just piling on hardware.
The staking mechanism is also very interesting. Provers must stake ZKC to participate, and the amount staked determines how much work you can take on. This is somewhat like taxi drivers needing to pay a deposit to drive. If the proofs you generate have issues, the staked tokens will be forfeited. This ensures the security of the network.
I looked at the profit model. Suppose you have 10 3090 cards, each card can generate 100 proofs per day, with an average profit of 0.1 ZKC per proof. At the current price of $0.3, that would be $300 a day, or $9000 a month. After deducting electricity and other costs, the net profit could be around $5000. This is more profitable than mining most altcoins.
However, there are several variables involved. First, there is the demand for proofs. If there are not enough applications using Boundless, provers will have no work. Secondly, there is competition. As more provers join, the cake will get smaller for everyone. Lastly, there is the price of the token. If ZKC drops, the returns will naturally decrease.
There is a detail worth noting. Boundless's proofs are processed in batches. Provers do not generate proofs one by one; instead, they package multiple requests together for processing. This greatly improves efficiency and reduces the cost of on-chain verification. This aggregation mechanism makes the entire system more economical.
The competitive landscape is also changing. In addition to Boundless, there are other projects doing similar things. For example, Succinct's prover network and some specialized ZK proof service providers. But Boundless's advantage is its openness. Anyone can become a prover without permission or KYC.
In the long run, I think proof generation will become an independent industry. Just like today's cloud computing, AWS and Alibaba Cloud provide computing resources that application developers use as needed. In the future, there may be specialized proof service providers offering various types of ZK proof services.
Of course, there are risks. The biggest risk is technological iteration. If a more efficient proof algorithm emerges, existing hardware may be eliminated. Additionally, there is the risk of centralization. If a few large players control most of the computing power, it could affect the decentralization of the network.
Overall, I think this is a direction worth paying attention to. Blockchain needs more computing resources, and idle GPUs can provide these resources. Boundless has created a market that allows supply and demand to match efficiently. If this model works, it could become an important part of Web3 infrastructure.
For friends with GPUs, I recommend giving it a try. Even if you don't go all in, testing with a few cards is also good. After all, if they're just sitting idle, it’s better to let them generate some value.