Topic: Coinbase: FACT0RN blockchain with Integer Factorization as PoW (Read 65 times)
Seems like a great project. Reminds me of Kaspa.
Perhaps you should read the whitepaper. That way you can avoid yelling, and perhaps, learn why gHash is designed the way it is. But, who knows? Maybe reading is not your forte. Incidentally, gHash does take a neglible amount of time while still protecting the blockchain from a mathematical attack using gHash that is described in the whitepaper. Again, reading the whitepaper will allow a conversation that is much more productive than the one we are currently having.
The whole point of gHash is to make it hard to implement in an ASIC
Your chain is called FACT0RN, not GHASH.
The whole point is to make a PoW dominated by factoring.
Any preparatory hashing should take a relatively negligible amount of time,
not so much that you have to start worrying whether it could be optimized by ASIC.
The whole point of gHash is to make it hard to implement in an ASIC, not to make it simpler. The complexity is exactly what is needed to make it hard to implement on ASICS, or FPGAs, for that matter. Plus, this is fast enough to verify quickly. The block target time is 30 minutes.
> gHash runs in about 5−10 milliseconds on x86; tested on both Intel and AMD processors.
Implementing this function on an ASIC would be a nightmare, and it is designed to be. The
resource consumption is vast in logic and memory
Making gHash super complex and hard to verify runs counter to the design principles for a good PoW.
If vast resource consumption is the goal, then this can be achieved without high complexity and slow verification.
E.g. replace gHash by Cuckoo Cycle [1], the simplest possible memory hard puzzle, or by Equihash.
[1] https://github.com/tromp/cuckoo
Implementing this function on an ASIC would be a nightmare, and it is designed to be. The
resource consumption is vast in logic and memory
Making gHash super complex and hard to verify runs counter to the design principles for a good PoW.
If vast resource consumption is the goal, then this can be achieved without high complexity and slow verification.
E.g. replace gHash by Cuckoo Cycle [1], the simplest possible memory hard puzzle, or by Equihash.
[1] https://github.com/tromp/cuckoo
Coinbase published an article(https://blog.coinbase.com/fact0rn-blockchain-integer-factorization-as-proof-of-work-pow-bc48c6f2100b) 3 days ago on Friday where one of the recipients of their Crypto Community Fund Awards (user https://twitter.com/LionesEscanor on twitter). (The website of the blockchain is https://fact0rn.io)
The whitepaper can be found here (https://fact0rn.io). Has anyone checked this out? Is anyone mining this stuff? What do you guys think?
The whitepaper can be found here (https://fact0rn.io). Has anyone checked this out? Is anyone mining this stuff? What do you guys think?
Jump to: