The highway analogy: Vitalik Buterin’s plan to scale Ethereum 1000x

Vitalik Buterin argues that working on increasing Ethereum’s bandwidth will potentially provide infinite scalability rather than working on reducing latency.

In a recent commentary on Ethereum’s future, Vitalik Buterin discussed a significant pivot point: the network should prioritize increasing bandwidth (capacity to process data) over reducing latency (speed of transaction processing).

His core argument rest about the difference between what is difficult and what is physically impossible. Reducing latency, which essentially means the time it takes to confirm a transaction, is fundamentally limited by the speed of light, among other things:

“We are fundamentally limited by the speed of light, and furthermore, we are also limited by […] the need to support nodes in rural environments, globally, and in home or commercial environments outside of data centers, the need to support censorship resistance and anonymity for nodes, etc… [Quote shortened for editorial purposes].”

Although it sounds complicated, there is a relatively simple way to visualize this concept.

The highway analogy: bandwidth versus latency

To best understand what Buterin is talking about, imagine Ethereum as a highway. The problem is traffic-related: you want to transport more people.

You really only have two options:

1. Reduce the latency (speed) of transportation by making each car move a lot faster.
2. Increase the bandwidth (capacity) of the highway by building more lanes.

Buterin argues that Ethereum should choose the second option because making cars go faster is dangerous and fundamentally limited.

This also ties in with another post he wrote five years ago, titled The limits to the scalability of Blockchain. In it, the co-founder of Ethereum discussed a tweet from Elon Musk:

You might also like:

Ideally, Doge speeds up block time 10x, increases block size 10x, and reduces costs 100x. Then it wins hands down.

— Elon Musk (@elonmusk) May 16, 2021

Vitalik outlined how challenging this is without sacrificing security and decentralization. Essentially, he argued that ordinary people should be able to use nodes on their standard equipment (such as laptops), meaning that the blockchain is obviously limited by the capabilities of this equipment. The alternative is for nodes to be managed through data centers, which entails the risk of centralization:

The elites of your blockchain community, including pools, block explorers, and hosted nodes, are probably quite well coordinated; most likely they are all in the same Telegram channels and WeChat groups. If they really want to If they want to organize a sudden change in protocol rules to further their own interests, they can probably do that.

That said, he also said that latency can indeed be reduced with existing technologies, without making compromises. These include:

• P2P (esp erasure coding) improvements can reduce message propagation times without requiring individual nodes to have lower bandwidth.
• An available chain with a smaller number of nodes per slot (e.g. 512 instead of 30,000), which can eliminate the need for an aggregation step, allowing the entire hot path to occur in one subnet.

According to Buterin, this could scale Ethereum alone three to six times and is “very within the realm of possibility.”

The world heartbeat

And yet, this specific technical difference defines Ethereum’s ultimate goal – to be the heartbeat of the world rather than the The world’s video game server.

He argues that with existing technologies such as PeerDAS and ZKPs, the means to scale exist and that Ethereum can scale thousands of times, at least compared to the current status quo. He also outlined that the figures will be a lot more favorable than before, and that there is no limitation that prevents ‘combining extreme scale with decentralization’.

Following my earlier analogy, this simply means that there is nothing stopping the team from building a highway with as many lanes (possibly thousands) as necessary.

However, this also means that there will come a time when applications require speeds faster than the heartbeat (e.g. Ethereum). In that case, Vitalik says these applications must have off-chain components. This also means that layer two scaling solutions will continue to play a role in the future, even as Ethereum scales massively.