Vitalik Buterin took to Twitter to announce the progress with Ethereum’s sharding design.
— vitalik.eth (@VitalikButerin) April 30, 2018
The tweet, which was expanded upon with a reddit post, outlined the design that will likely be implemented on the Ethereum blockchain.
Sharding is Ethereum’s answer to the scalability problem. Currently, the network can handle around twenty-five transactions a second. As more users participate, a backlog of transactions needing to be mined can build up, slowing the network to a halt. In comparison, Visa can handle around forty-five thousand transactions per second – to match this, the Ethereum blockchain needs to fundamentally change.
Sharding, the proposed solution, will see the blockchain split from one chain of blocks to multiple chains working parallel. This model will massively increase the throughput, but details had not been released explaining how the system would function.
The update from Vitalik, creator of Ethereum, sheds some light. The proposal will revolve around a beacon chain, which will act as a heartbeat for all the sharded chains and the mainchain. By using randomness, the beacon chain will coordinate the sharded chains, keeping the entire system secure and in order.
The first iteration of this model allows for twelve sharded blockchains, but in the final implementation Vitalik expects the network to be able to handle one hundred. This upgrade would massively boost the capabilities of the network, allowing it to support the huge ecosystem of decentralized apps building on the network.
An interesting feature of the proposed system is that the capability of the network would scale exponentially, as Vitalik explained:
Yes. And throughput increases from there will be quadratic, ie. if computers get 2x as powerful, the blockchain’s theoretical max capacity will increase by 4x.
Whilst sharding is still in early stages, Ethereum’s scalability upgrade is well underway.