What is ZK-IBC / ZK-tendermint?

ZK-IBC, (or ZK tendermint), refers to the ZK Light Client for tendermint consensus. As explained in this previous article by Electron, we can use ZK-proofs to create cheap proofs of consensus validity.

Ever since we published this article, many companies have that come up with various implementations of ZK Light Client for tendermint, including but not limited to Electron, Polymer, Union, Succinct, among others.

Challenging with existing ZK-IBC implementations

However, inspite of this, ZK-IBC remains impractical today due to the very high gas cost of verifying ZK-proofs on Ethereum. For eg, verifying a single zk light header costs >250K gas. So, for Osmosis with a blocktime of 5s, and Eth blocktime of ~12s, we will need to submit ~2 proofs per Eth block. At a gas price of 60 Gwei, this costs $100 per 12s OR $260mn per year, just to keep the Osmosis light client in sync. This is absolutely not feasible. Note that this cost will only increase with time as Ethereum gets more expensive.

In order to solve this problem, we are now launching the next generation of ZK-IBC technology that will reduce the gas cost by orders of magnitude, and make ZK-IBC great again.

What is ZK-IBC 2.0?

Our Solution is simple - we aggregate individual zk-proofs from multiple chains into a single "superproof". This enables the cost of proof verification to be amortised across many chains. The current state of the art ZK-tech already enables us to aggregate a large number of cosmos chains, with reasonable proving costs. The testnet for ZK-IBC2.0 is live.

Measuring Cost Improvements

Assuming a proof verification gas cost of 250K, and N Cosmos chains being aggreagted, the cost per chain is 250K/N. For N =10, the cost per chain is 25,000, which is roughly the cost of an Eth transfer. As we add more and more chains into the network, the proof verification cost per chain continues to reduce and tends to zero, till the point the zk-hardware cost dominates.

Therefore, we can achieve multiple order of magnitude cost reduction compared to current levels.

Over time, ZK-IBC 2.0 will get cheaper than even centralized bridges.

Last updated