Data shows it's not as cheap as imagined! Vitalik: The current rollup Layer 2 solution is far from being optimized

share
Data shows it

On 7/29, Ethereum Layer 2 scaling solution Optimism reached a historic high of $550 million in on-chain assets, making it the second largest in terms of funds among current rollup scaling solutions. With the anticipated Ethereum merge news, Optimism's governance token OP also surged nearly 200% in 30 days, indicating market optimism for the fast and low-cost environment that L2 is expected to bring.

Rollup Is Not as Cheap as Expected

Sanjay Shah stated that there is a gap between the theoretical and actual network fees of the rollup scaling solution. Here is his research:

Firstly, to understand, rollups batch process transaction data and periodically record them on the L1 Ethereum mainnet. The Gas spent on L1 can be split into two parts: a fixed Gas cost for each batch and a variable Gas cost for each transaction.

1. Optimism Optimistic Rollup

Sanjay Shah found that the current batch Gas is around 280k, and each transaction Gas is about 3,300. However, this differs significantly from Vitalik's estimation in 2021, where he predicted the batch Gas to be 40k.

2. Arbitrum Optimistic Rollup

Sanjay Shah found that in another rollup solution, Arbitrum, with $840 million in on-chain assets, the required batch Gas is around 375k, ten times more than Vitalik's estimation.

Sanjay Shah believes that the lower batch fees are one of the aspects where optimistic rollup outperforms zk rollup, another commonly mentioned technical approach. However, the above data indicates that it may not stand out as originally thought.

3. zkSync zk Rollup

Sanjay Shah's observation of zkSync data shows that the batch Gas is around 800k, and each transaction Gas is about 600. This is closer to Vitalik's prediction of 500k batch Gas, but he is still surprised by the transaction costs.

According to Polygon's zk rollup solution, each transaction should only require 14 bytes, and each byte of calldata needs 16 Gas, which is 14x16=224, meaning each transaction should only need 224 Gas. Sanjay Shah is curious why zkSync requires such high fees, suggesting that zkSync 1.0 may not be fully optimized and could see cost reductions in zkSync 2.0.

Data Source: Sanjay Shah's Data Analysis