# Meeting 29/07
## Potential candidates
### Ethash
* coin: Ethereum or Ethereum Classic
* Very popular
* Well documented
* Memory hard (>16 MB cache)
* Python implementation is available
* The [ProgPOW](https://github.com/ifdefelse/ProgPOW) page claims Ethash can be accelerated to double speed with an ASIC
* Ethereum is planning to move to a proof of stake algorithm in the [near future](https://cointelegraph.com/news/eth-miners-will-have-little-choice-once-ethereum-20-launches-with-pos)

### Beamhash III
* Coin: Beam
* Open source
* Based on Equihash/ Blake 2B and SipHash 2,4
* final PoW algorithm -> stability
* Designed to transition to ASICs
* but still single chip ASIC resistant for the next year(s)
* Only 1 month old (end of june 2020)
* C code available

### Cryptonight
* Coin: Haven, Bloc, Ryo...
* Open source
* memory-sensitive algorithms
* Based on well known hashes (AES, SHA3 competition finalists)
* different versions (lite, heavy, 'original')
* Some versions are already implemented on ASICs so maybe less interesting

### Cuckoo cycle
* Coin: Grin, Aeternity
* Open source
* Calculate solution to a graph theory problem
* memory-bound algo -> less energy than most other PoW algo's
* ASIC-friendly (cuckatoo) as well as ASIC-unfriendly (cuckaroo) versions
* Bounties for more efficient implementations are available

### neoscrypt

[FPGA](https://fpga.guide/en/downloads/)

## Conclusion

Look into implementations of CryptoNight, specifically Haven or some other popular variant.

If none are found, we go with BeamHash III.