What factor determines speed?

[graphenepay]

Comparing my 'regular' PC (16cores) against my server (96 cores), there is almost 0 speed profit.
What is the decisive factor to run on-chain tests at faster speed?

[shouc]

Higher CPU frequency and faster node access give better performance.

[graphenepay]

And the etherscan integration for onchain? Is that needed? I assume it limits the execution speed as well

[shouc]

You can build with no_etherscan feature. ItyFuzz would then use the decompiler to replace Etherscan.

cargo build --release --features no_etherscan