From 2015, IOHK sought how to make Proof Of Stake (POS) practical. They had to look at the entire science of consensus protocols for the crypto industry and what was required was a practical solution to gradually scale as users entered the system. By contrast for most existing protocols, as more users are added, the resources are stretched and do not increase, thus the network suffers from congestion, higher fees, slower resolutions and a generally bloated experience.
The objective was to build a system that when it becomes more decentralized, the network performs the same or ideally better than a less decentralized state. A functioning example of this kind of decentralized increase of network performance is BitTorrent, whereas the more users there are hosting and sharing a file, it becomes easier and faster to continue to share that file.
This was an enormous ask with so many considerations.
“Adaptive security, semi-synchrony, we wanted fast finality, we wanted the ability to bootstrap from genesis without having a checkpoint, we wanted to de-couple the clock. We wanted to be able to recover from spikes of dishonest majority and also we needed to build a delegation system, we needed to build a stake pool system and that we needed a productive way of orchestrating, curating all of that including building a brand new way of handling networking for our entire industry.”
This colossal task has consumed 5 years of research from a team of more than two dozen scientists and engineers essentially working full time.
As a result of this ask and these efforts is what is called
The isomorphic state channel paper can be viewed here on eprint.
Ouroboros Hydra is a layer 2 solution that can sit gracefully on top of the Cardano protocol and accelerate it. Simulations show that each Ouroboros head can perform around 1000 transactions per second (TPS) (un-optomized). As more “heads” are added to the protocol, these heads should show a similar TPS.
“For example if you have 1000 TPS and 1000 heads [stake pools] you could be looking a maximum theoretical performance of a million transactions per second.”
These heads could offer fast finality and micro transactions (micro-payements, micro-tipping).
“This paper is the first interdisciplinary paper where we had network people, NPC people, programming language people, and consensus people working together to figure out the theory and we’re so excited that the smart contracts that can run on the base layer we have shown are isomorphic with what can run in the state channels. Meaning that if you have a Plutus smart contract in our model, it’s going be able to run in that head 1 for 1. This is not a trivial thing to show and it’s not a trivial thing to do.”
Implications of Ouroboros Hydra
“What is most exciting about this is that it can be implemented in parallel with all of our work on Shelley Goguen and Voltaire.”
“This solution is complimentary to ledger based sharding solutions that we’re seeing in industry.”