Multiple proof-of-work algorhytms simultaneously

There is an idea of one coin.. Myriadcoin. Good idea IMO !

Theoretically - it is possible to build a coin, that has several algos, and switches between them.
If we have 2 algo:
1. SHA256
2. SHA3

Then even blocks could only be mined using first algo, and uneven blocks could only be mined using the second algo, making ASICs more difficult to build.

What do you think of it ?

-Technologov, 08.Apr.2014.


  • troytroy Member Posts: 9
    Makes sense... Good idea... IMO.
  • VitalikButerinVitalikButerin Administrator Posts: 84 admin
    Problem is, that doesn't really solve the problem, it just delays it. Instead of having to ASIC one algo, you have to ASIC ten algos. Eventually it will get done, and all it really does is reduces the multiplier from maybe 10000x to 1000-3000x. The idea of multiple algos is definitely a good way to go, so we could turn it into an optimal algorithm by bringing in an idea from Ethereum: the algo constitutes running contracts, allowing any new mining algorithm to be specified by adding in a new contract, and the system is set up so that anyone, perhaps meeting certain conditions, is free to add in a new algo. Thus, if an ASIC manufaturer does appear, everyone else has the incentive to deliberately create a contract that they cannot mine efficiently and "poison the well". An individual nonce would randomly select at least 16 contracts, so it is not practical to "shop around" for the single algo you're best at. The one weakness of such a system is that it's possible to create algos that everyone else is bad at but you're good at (eg. find a sub-nonce that RSA decrypts to the given nonce under your public key), and so we may end up seeing the blockchain inundated by such deliberate bad algorithms. If we can figure out a good solution around this, I think it would be close to optimal.
  • TechnologovTechnologov Member Posts: 102 ✭✭
    In a related note, I have found a possibility to encourage more people to run full nodes. The trick is Proof-of-Stake plus demurrage of inactive keys.

    See Cunicula's Implementation.
  • TechnologovTechnologov Member Posts: 102 ✭✭
    I hint at mixing-in Proof of Stake algo.
  • JasperJasper Eindhoven, the NetherlandsMember Posts: 514 ✭✭✭
    Was under the impression proof of stake as addition is on the table already.
Sign In or Register to comment.