Sharding may return to scale Ethereum to millions of TPS

Update (Nov. 12, 8:47 am UTC): This article has been updated with additional comments from Joe Lubin.

An Ethereum researcher’s social media post has ignited speculation about a possible solution for the layer-1 blockchain’s scalability challenges. Consensys CEO Joe Lubin said this may involve reviving the concept of sharding to address scalability.

On Nov. 11, Ethereum researcher Justin Drake posted on X that he would announce an “ambitious” initiative for Ethereum. Drake said he had contemplated a “from-scratch” redesign of the Ethereum consensus layer, which some interpreted as a step toward solving its scalability issues.

The researcher said his goal was to suggest a strategy to ship a Beacon Chain road map. He is expected to share the full proposal at Devcon in Bangkok, Thailand, on Nov. 12.

Bringing back sharding to Ethereum

In an interview with Cointelegraph’s Andrew Fenton, Lubin suggested Drake’s ideas involve revisiting sharding execution. The CEO explained that this could entail developing a zero-knowledge Ethereum Virtual Machine (zkEVM) at layer 1, allowing for multiple identical execution shards. 

Lubin explained that while sharding was set aside a few years ago because it “wasn’t really possible,” it might make a comeback in the upcoming proposal: 

“The interesting thing about that, that way of using layer 1 wasn’t really possible a few years ago when we discarded the idea of execution sharding, what we needed to do was throw open this divergent exploration and a lot of stuff came back.”

Lubin added that there’s a lot of learnings from the development of zero-knowledge approaches and optimistic approaches that could be brought back down to the Ethereum layer-1 and “make everything better.”

Lubin also believes that this could also lead to scalability solutions for Ethereum. “You’re just boiling down a giant amount of computation at different layers and amortizing a lot of computation into a single transaction. If you do that every two seconds or less, then you get a lot of transactions per second,” he explained. 

While Lubin anticipates millions of transactions per second (TPS), he also cautions that achieving this scale could take years.

Related: Consensys to cut workforce by 20% — CEO outlines decentralization plans

Community speculates about ETH 3.0

Following Drake’s post, rumors about an ETH 3.0 upgrade circulated among the Ethereum community. On X, Ambient Finance founder Doug Colkitt posted about a rumor that the ETH 3.0 announcement is a “second merge into a new consensus targeting 1-second block times” and a native zero-knowledge Ethereum Virtual Machine (zkEVM). 

Colkitt said that if the rumors turn out to be accurate, having a native zkEVM would be a “huge” update:

“The gas limit can be eliminated entirely. Builders can build arbitrarily large blocks since nodes only need to verify the snark. The only scaling limit left would be bandwidth.”

Colkitt expressed optimism that a zkEVM may mean arbitrary scalability and eliminate the need for layer-2 rollups.

Not everyone in the community was on board with the ETH 3.0 speculation. One community member said the rumor was “100% BS,” pointing out that such significant updates would have been signaled months in advance. The community member noted that related Ethereum Improvement Proposals would likely have been filed if such an update were imminent. 

Magazine: We took an ETHSafari to see how crypto is working out in Africa