Cryptoabc用户3885
Week 7 Research Track - Verkle Trees
输入“/”快速插入内容
Week 7 Research Track - Verkle Trees
Week 7 Research Track - Verkle Trees
用户3885用户38852024年4月7日修改
Guest speaker
Summary notes
•
•
EF stateless consensus team & Website info
•
Team member
•
Motivation of the Verkle
•
Stateful applications are complex
◦
State in systems creates many challenges, eg. state access, efficiency, storage
◦
State (usually) only grows with time
•
For Ethereum, it puts some pressure on core values
◦
Currently, to be able to validate blocks, the prerequisites incl.
▪
Download all the state, which takes time
▪
Save the state somewhere -> a full note requires ~ 1T + 300GB disk
◦
Handling state isn't zk-friendly
•
Towards a better stateless world
◦
A new node can join the network
▪
Without the need to sync all the state
▪
No requirement of disk storage for the EL client
◦
Become Zk-friendly L1, by
▪
Removing complex data structure (MPT)
▪
Removing heavy use of non-zk-friendly hashes (i.e. Keccak)
◦
Reduce hardware requirements
◦
Simplify the implementation of a new (stateless) EL client
◦
Potentially allowing increasing the gas limit
◦
Might trigger the specialization of protocol roles, eg. different nodes have different roles in the protocol
Design & Architecture
Tldr of the Verkle
•
Execution witness: Key concept introduced into the protocol, which
◦
Contains all the state needed to execute a block
◦
Contains a very small cryptographic proof to validate the correctness of the witness
◦
Contains all the code needed to excute the contracts
•
The execution of Verkle transition is complicated
◦
Need to introduce a new cryptography stack
◦
Need to change the state tree(s) data structure
◦
Need to change gas accounting
◦
Need to migrate data from MPT to VKT