An Ethereum developer, Georgios Konstantopoulos, has shared the details of the Berlin network upgrade. In a series of tweets, the developer stressed the importance of the upgrade to the Ethereum network. He stated that the upgrade is scheduled to take place on mainnet in block 12,244,000, about 41 days’ time.
In the thread of tweets, the Greek developer also shared the necessary things that the Ethereum community members need to know about the changes in the accepted Ethereum Improvement Proposals (EIPs), as they anticipate for Berlin hard fork.
Ethereum Improvement Proposal (EIP) 2929
Regarding the EIP-2929, Konstantopoulos stated that it “increases the gas cost for the first time a slot or account is accessed.”
Adding that it also “reprices the “cold” cost of account access opcodes (BALANCE, *CALL, EXT*) to 2600, and reprices state access costs (SLOAD) from 800 to 2100, while setting the “warm” access costs to 100.”
According to his further illustration, the first time an account or storage slot is accessed, users pay more, but the cost will subsequently reduce at further accesses.
“It also adjusts cold SSTOREs to charge an additional 2100 gas which can be refunded to the user according to the net gas metering rules defined in EIP-2200. This is needed to avoid a DOS attack, and actually makes SLOAD->SSTORE cheaper than before.”
Ethereum Improvement Proposal (EIP) 2930
In the case of EIP-2930, the developer said it’s designed to reduce state accesses, adding that it “allows warming-up an account (1900 gas) or storage slot (2400 gas), meaning instead of paying 2100 per follow-up SLOAD or 2600 per EXT*, you’d pay 100!”
According to Konstantopoulos, EIP-2930 is also a great step towards stateless Ethereum, helping with its long-term scalability at the protocol layer. However, access lists need to be introduced in a backward-compatible way, which is why EIP-2718 is introduced.
Ethereum Improvement Proposal (EIP) 2718
Regarding EIP-2718, known as Typed Transaction Envelope, the Ethereum developer said it defines a new transaction type, which is an envelope for future transaction types.
Konstantopoulos wrote, “EIP-2718 is important because it’ll allow an easy transition from the current transaction format to access-list-enabled transactions. Clients will just inspect the first byte and know the tx type by that. It’ll also make it easy to enable EIP-1559 txs, without code churn.”
Ethereum Improvement Proposal (EIP) 2565
EIP-2565, known as ModExp Gas Cost, implements modular exponentiation, which is an important operation for building cryptographic primitives such as SNARKs, VDFs, RSA Accumulators, and more, said Konstantopoulos.
He further stated that Modexp opcode costs ~5-20x more than it should, so the EIP-2565 changes the gas cost calculation formula to tune down the cost, making on-chain crypto transactions cheaper.
Ethereum Improvement Proposal (EIP) 2315
The last Ethereum Improvement Proposal as listed by the developer, Georgios Konstantopoulos, is EIP-2315, which is referred to as Simple Subroutines for the EVM.
According to Konstantopoulos, “Subroutines in the EVM are not directly supported. This EIP adds a new stack/opcodes to store for ret pointers, improving bytecode static analysis and with marginal gas cost improvements.”
The illustration shared above is the summary of the changes in the upcoming Ethereum hard fork, known as Berlin network upgrade. And as aforementioned, the hard fork will take place in block 12,244,000. Mind you, this precedes London hard fork that is slated for July 2021.