Oracle Data Feeds on zkRollups are a complete game changer economics wise
A Data Feed with hundreds of nodes posting thousands of updates per batch has the same L1 gas cost footprint as a price feed with 1 node posting 1 update per batch
A Data Feed with hundreds of nodes posting thousands of updates per batch has the same L1 gas cost footprint as a price feed with 1 node posting 1 update per batch
How?
zkRollups like @dydxprotocol don't post the raw tx data on-chain, rather they post compressed state diffs (state differences from batch n and batch n+1)
A Data Feed is an on-chain reference contract with a variable that holds an aggregated value that's updated by oracles
zkRollups like @dydxprotocol don't post the raw tx data on-chain, rather they post compressed state diffs (state differences from batch n and batch n+1)
A Data Feed is an on-chain reference contract with a variable that holds an aggregated value that's updated by oracles
So no matter how many oracle nodes or updates there are, there will only ever be at most 1 state diff per batch for a Data Feed, since updates always touch the same state variable
Signature verification and data aggregation is succinctly proven within the validity proof
Signature verification and data aggregation is succinctly proven within the validity proof
The most practical impact from that is that the update frequency of Data Feeds on a zkRollup can be much higher, fresher data at a lower cost for users
Increasing the number of oracle nodes should be taken with care given not all oracle nodes that exist are of the same quality
Increasing the number of oracle nodes should be taken with care given not all oracle nodes that exist are of the same quality
• • •
Missing some Tweet in this thread? You can try to
force a refresh
