Layer 1 Proof-of-Work (PoW) has made big advances in blockchain research, but not in implementations, mostly due to Bitcoin’s reluctance to upgrade its protocol. Ergo is in a prime position to leverage state-of-the-art PoW research and fully cement its place as the ultimate PoW blockchain. In this forum post I introduce Prism and why it’s a good fit for Ergo as a long term Layer 1 goal.
Researchers at Stanford have developed the first PoW consensus protocol scaling up to network throughput with fast finality. The core logic stays very similar to Ergo’s PoW we know and love, and doesn’t sacrifice decentralization to achieve better performance.
Core idea
Instead of having a single chain handle adding transactions and voting for the main chain, Prism separates these functions into separate types of blocks:
- Transaction blocks, contain the transactions and can be referenced by their short header.
- Proposer blocks, links to transaction blocks and is the equivalent of regular PoW’s main chain
- Voter blocks, vote on the canonical proposer blocks to achieve consensus.
The main chain of proposer blocks is decided by the vote of multiple voter chains, each run by longest-chain consensus. Votes by voter blocks become confirmed as their depth in their respective chain increases, same as a regular PoW chain. With all the voter chains combined, Confirmation on the proposer chain is achieved much faster than by waiting for a single chain to achieve confirmation.
Miners can randomly mint either of these blocks and broadcast them. The frequency of proposer blocks would be the same as Ergo’s current block frequency, while transaction and voter blocks could be more frequent depending on the current needs of the chain.
I recommend understanding more details with this well explained presentation by one of the authors Dr. Giulia Fanti: MIT Bitcoin Expo 2019 - Deconstructing the Blockchain to Approach Physical Limits - YouTube
Full paper: https://dl.acm.org/doi/pdf/10.1145/3319535.3363213
Bonus smooth mining rewards benefit
In the basic PoW protocol, miners are able to mint a block when they find a hash value under the difficulty threshold d, let’s say in the range [0, d]. Other types of blocks can be minted in other ranges, [d, 2d] or [2d, 3d] etc… This means it is easier to mint “something”, although probably with lower rewards per block to keep the emission schedule on track. The added benefit is the better viability of solo mining, as less hash power is needed to mint “something” from time to time.
Step by Step roadmap
Prism doesn’t have to be implemented all at once thanks to its structure. Just adding transaction blocks and having the main chain link to transaction blocks instead of including the transactions itself would already greatly increase scalability and offer miners more chances at rewards.
The voter chains system is more complex and actually just brings fast confirmation, which might not be necessary for Ergo depending on Layer 2s’ progress