zkPorter Promises to Bring 20,000 TPS to Ethereum

zkSync introduced an Ethereum Layer 2 solution called zkPorter yesterday promising to push the blockchain’s transactions per second (TPS) capabilities to over 20,000. This would put Ethereum within striking distance of Visa, which says its network can handle 24,000 TPS.

“To bring true scalability, we need an exponential gain in throughput,” the team wrote in a blog post. “That’s why we believe this is such an important breakthrough. We have designed a system with 20,000+ TPS that offers more security than optimistic rollups. More importantly, it is coming to mainnet in 6 months with zkSync 2.0.”

zkPorter accomplishes its high TPS capabilities by publishing its accounts’ data off-chain. Holders of the zkSync token, which is not yet live, will secure zkPorter accounts’ state with the Proof of Stake (PoS) mechanism. As the launch post says “the majority of rollup fees are due to the cost of publishing data on Ethereum.”

According to the team, zkPorter will be fully interoperable with zkSync’s zkRollups when the project’s 2.0 iteration goes live on Ethereum mainnet in six months.

The zkRollup accounts are ultra-secure as they have the same security guarantees that Ehtereum’s Layer 1 does, but the zkPorter side will be less secure as it will depend on the zkSync holders, whose token will almost certainly have a much smaller market capitalization than Ethereum, making it more susceptible to attacks. The zkSync team stresses however, that, unlike optimistic rollups, where successful attacks allow for a rollups’ funds being stolen, zkPorter attacks could only halt block production.

The increase in TPS capability means the price of single transactions will drop as users won’t need to bid high amounts of gas for their transactions to be processed. zkSync promises that for users with zkPorter accounts, Uniswap trades will cost below $.03 which is over 100 times cheaper than the transaction currently costs on Layer 1.

Vitalik Pushes Back

On Reddit, Ethereum co-founder Vitalik Buterin pushed back on the zkSync team’s claim that zkPorter offers better security guarantees than optimistic rollups, saying that the approximation of an attack on the latter costing $50M was unrealistically low because the hashpower needed will become more expensive as a party rents it with the intent to attack.

Buterin also believes that, in the case of an attack on an optimistic rollup, the Ethereum community can easily coordinate a soft fork. In the case of such an attack, the Ethereum community will have one week to coordinate a soft fork before the rollup’s data is submitted to Layer 1. Buterin cited an Ethereum hard fork, which requires all nodes to upgrade to the agreed on blockchain state, rather than just the simple majority as with a soft fork, which was coordinated in six days as evidence for the communities’ ability to act.

Buterin also believes that the data availability of zkPorter accounts will be much less secure as it will depend only on zkSync stakers, rather than the entire Ethereum network.

A zkSync team member replied to Buterin’s concerns, saying that their point is that zkPorter offers similar levels of security guarantees as optimistic rollups, but are much cheaper. “Why would anyone need optimistic rollups?” they asked.

The poster went on to acknowledge that in the “long term ETH 2.0 and 3.0 will rule,” referring to the planned updates to Ethereum which will bring scaling to the chains layer one by running many PoS chains in parallel, a process called sharding.