Monad, an EVM-compatible blockchain, activated its MonadTen revision on mainnet at 14:30 UTC on Sept. 2. The MIP-8 upgrade changes how the network charges contracts for reading storage, replacing slot-by-slot warming with groups of 128 consecutive slots.
Monad’s mainnet block 101672712 measured 8,100 gas to read slot 0, 100 gas to read slot 1 on the same page, and 8,100 gas again at slot 128, the first slot of the next page.
The official MonadTen release record sets activation at Unix timestamp 1788359400. Below that timestamp, MonadNine rules remained in force.
Storage layout now changes the bill
The final MIP-8 specification defines a storage page as 128 words of 32 bytes each, or 4,096 bytes. The first SLOAD anywhere on a page costs 8,100 gas. Later reads anywhere inside that warmed page cost 100 gas for the rest of the transaction, subject to rollback when a call frame reverts.
Under Monad’s prior schedule, two previously untouched slots each cost 8,100 gas even when they sat next to each other. Under the new opcode pricing, the same pair costs 8,100 gas and then 100 gas if both fall within one page.
Common Solidity layouts can inherit that discount automatically. Sequential state variables, struct fields, and array elements occupy consecutive slots, so repeated reads are more likely to stay inside a warmed page.
A mapping key generally resolves to a separate, dispersed page, but fields inside a struct stored under that key remain contiguous and can share the page-level discount.
Hashed or unaligned storage keeps Monad’s existing cold baseline. Reads that cross page boundaries still cost 8,100 gas each, while savings depend on how many slots a transaction touches and whether those slots cluster within the same 128-slot boundary.
MIP-8 preserves EVM execution semantics while changing the assumptions used by tooling that builds access lists, storage proofs, or gas estimates. EIP-2930 entries now warm pages, and proof formats must represent the page model. The specification identifies contracts that hardcode storage-opcode gas costs as the main compatibility-risk class.
For developers, the incentive is that data read together is cheaper when it is stored close together.
