Excellent article. I hope Cuckoo ASICS will not be too difficult to make. Clearly more difficult than pure hashing PoWs though
The article speaks in favor of grin, and other cryptos that dominate their POW.
A future (next year presumably) grin where ASICs dominate graphrate.
I think it speaks in favor of Cuckoo too, as it is (if I am not mistaken) intended to be computationally simple.
Quoting from GitHub - tromp/cuckoo: a memory-bound graph-theoretic proof-of-work system :
“The memory efficient miner uses 1 bit per edge and 1 bit per node in one partition (or 1 bit per 2^k nodes with linear slowdown). It is bottlenecked by accessing node bits completely at random, making it memory latency bound. The core of this miner, where over 99% of time is spent, is also relatively simple.”
As far as memory intensive PoWs go, I don’t think you can get any simpler than that…
I am under the opinion that ASIC is not a problem but a solution to the security of a PoW-based network. Cheap access to ASICs is an equalizer. Given the circumstance where new ASIC models are rolled out over time, we should expect the first-to-acquire (manufactures or other small interests) to benefit the most; liken to a trickle down economic process. If innovations in ASIC hardware are negligible for ROI, we should assume lesser advantages in this way. Ideally, there should be no license or regulation enforced on the participation of mining: this goes for legislature and for the market.
Furthermore, I’m curious why VDFs aren’t explored more as a means for PoW. There is this property of VDFs where work is done synchronously and parallel work is rendered non-advantageous. I see this as helping in efforts to diminish returns on ASIC development. VDFs also share the property that of block aggregation: the dependency on previous block hashes in an aggregate of blocks is essentially a VDF and is the property which enables blockchain security.