The next Ethereum fork (Glamsterdam) is gold mine for bounty hunters. Only on the execution side: validator and builder modes, with/without BAL validation, parallel execution is mandatory. All these compound.
Fortunately, we are not shipping any serious EVM features.
This next fork is a high stakes one. Lots of changes that enable for a x5 increase in gas limit, a x4 in blob data, faster state updates, trustless slot selling, etc. At the same time it's so complicated that people involved have their nerves wrecked. Shipping during AI times
@eth_proofs@yukonresearch The challenge is flawed and has been exploited already. The cheapest solution is to return precomputed results for test vectors and fallback to proper implementation only for correctness.
https://t.co/hn4lSJLfQf is now live on @yukonresearch
Join the open autoresearch challenge to write and benchmark EVM bytecode replacements for EIP-8200 precompiles, establishing proving cost baselines for zkVM provers and helping simplify the protocol.
Each promoted submission becomes the shared record to beat. Enter your agent ↓
https://t.co/SUK0OR5dDG
Why is EVM bytecode so hard to decompile? Dynamic jumps, plus compilers reusing the same instruction blocks across unrelated functions. Yannis Smaragdakis on the core problem, in under a minute.
https://t.co/OIzT5MHFbd
ZisK v1.1.0-alpha release 🚀
Ethereum proven in real time on 4x5090 GPUs with a p99 of 9.62s. All providing 128-bit provable security, powered by Poseidon1, and post-quantum secure.
This release pushes the prover further and fixes bugs, without trading soundness for benchmarks.
Release notes: https://t.co/J8tSAalTMi