Paper weighs one fee price against separate prices for Ethereum

A paper on one meter versus many

Fei Wu has published a paper comparing fee mechanisms on Ethereum that rely on a single price against those that set separate prices for different resources. The post appeared on the Ethereum Research forum on 12 November 2026.

The title is “More Meters or More Prices? One- vs. Multi-dimensional Fee Mechanisms on Ethereum”. According to the post, the research builds on earlier work. Previous posts analysed the equilibrium and dynamic simulation of the multidimensional fee market proposed in EIP-7999, and compared it with the one-dimensional market that is expected after an upgrade the post calls Glamsterdam.

Wu writes that part of the work was done during an internship at the Ethereum Foundation and thanks @misilva73 for comments and discussion.

The normalised benchmark

The paper evaluates three one-dimensional designs as benchmarks, according to the post. The first is a baseline with a floor of 64 gas per byte and a parameter Wu calls CPSB equal to 1530.

The other two build on that baseline. One pairs a floor-adjusted design with EIP-8368. The second pairs it with EIP-8372.

The post states that, under current resource demand, the strongest of the three is the combination with EIP-8372. The paper calls this the normalised benchmark.

In this design, the transaction floor price is adjusted to each propagation allocation. EIP-8372 also scales CPSB and the state gas limit by a common factor. The stated aim is for regular gas and state gas to reach their targets at the same base fee in equilibrium. The design keeps the target-to-limit ratio of 1/2 and the update rule from EIP-1559, according to the post.

Block-level demand variation

The excerpt describes results under variation in block-level demand. In that scenario, the post says the design delivers 151.9M to 177.9M execution gas per block.

The excerpt ends there. The post does not, in the material available, set out how the multidimensional mechanism performed against the normalised benchmark. It also does not state whether the paper settles the title question in favour of more meters or more prices.

Terms in the post are used without full definitions. CPSB appears as a per-byte gas parameter that EIP-8372 would scale alongside the state gas limit. The post refers to state gas and regular gas as separate resource types without explaining either.