4 ms·
For dedicated ASICs: It's still working, but the hardware can only perform the SHA-256 hash algorithm crazy fast, that's it. > it would be better to use a new
by zsmi 5y ago
For dedicated ASICs: It's still working, but the hardware can only perform the SHA-256 hash algorithm crazy fast, that's it.
> it would be better to use a new POW algorithm, which does not benefit from single purpose ASICs.
It's hard for me to imagine what that would be but that could be a lack of imagination on my part. Maybe other non-payment blockchain applications?
The other day some friends and I were musing how cool it would be if we could map some simulation algorithms to SHA-256 primitives. I thought there could be chance because I didn't actually know the mathematical operation SHA was doing. Once I found out, it didn't work out at all. Not even close. Maybe others will be more fortunate.
And for the one-off needs more mainstream CPUs already have support for some acceleration of SHA-256 [1] and that seems to be enough for them.
[1] https://en.wikipedia.org/wiki/Intel_SHA_extensions https://en.wikipedia.org/wiki/Intel_SHA_extensions
- tiluha 5y agoMonero uses RandomX as its hash algorithm, which is based on random code excecution and optimised for CPUs. https://github.com/tevador/RandomX https://github.com/tevador/RandomX
- tromp 5y agoAs I pointed out in https://news.ycombinator.com/item?id=28521804 https://news.ycombinator.com/item?id=28521804, hashcash with RandomX makes for a rather poor Proof of Work.
- tiluha 5y agoPlease tell me if i am wrong because i am not sure if it is right. Verifying 1 block requires the calculation of 1 hash. The hashrate for my CPU for RandomX is ~8000hashes/s. The monero blockchain currently has 2450000 blocks. blocks / hashrate = 5 minutes to verify the whole blockchain. Is that really that bad?
- tromp 5y agoMaybe you have a particularly beefy CPU. What if you have the original Raspberry Pi? Why do you think they made not checking the PoW supporting the checkpoint the default?