pith. sign in

arxiv: 2606.22206 · v1 · pith:LSHYJCDEnew · submitted 2026-06-20 · 💱 q-fin.TR

Transaction Costs and Speed in the Ethereum Ecosystem: Scalability of the Mainnet and Layer 2s

Meghan Ambrosia , Bruce Mizrach This is my paper

Pith reviewed 2026-06-26 10:33 UTC · model grok-4.3

classification 💱 q-fin.TR
keywords EthereumLayer 2transaction feesTPSscalabilitySolanablob expansionmedian fees
0
0 comments X

The pith

Ethereum L2 networks will exceed Solana TPS by March 2029 and match fees by October 2026, while mainnet fees converge in 2027 but TPS stays below 100 until 2034.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The paper examines transaction speeds and median fees on Ethereum mainnet, its Layer 2 networks, Solana, and Polygon from January 2024 through March 2026. Upgrades that increased block size and blob counts doubled TPS on both mainnet and L2s while cutting mainnet median fees from over $2 to under $0.02 and L2 fees by more than 95 percent. The authors project these improvement rates forward under continued blob expansion. They conclude L2s will surpass Solana on both speed and cost within the next few years, mainnet will match Solana fees by August 2027, and mainnet TPS will lag until 2034 even under the L1 Strawmap proposal. These timelines matter for understanding which parts of the Ethereum ecosystem can compete on throughput and cost with rival chains.

Core claim

The paper claims that with continued blob expansion, L2s will surpass Solana TPS in March 2029 and have lower median fees by October 2026. Mainnet median fees will converge with Solana in August 2027, but TPS will remain below 100 until 2034. The L1 Strawmap, proposing EIP-7938, brings the mainnet to only 100 TPS in January 2028.

What carries the argument

Linear extrapolation of TPS doubling and fee reductions observed after recent increases in block size and blob count.

If this is right

  • L2 median fees fall below Solana by October 2026.
  • Mainnet median fees converge with Solana in August 2027.
  • L2 TPS exceeds Solana by March 2029 under continued blob expansion.
  • Mainnet TPS remains below 100 until 2034, reaching only 100 TPS in January 2028 even with the L1 Strawmap proposal.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • Sustained L2 advantages could shift developer and user activity away from mainnet even if mainnet fees improve.
  • Polygon's inclusion in the comparison suggests the authors see room for multiple scaling paths within the broader ecosystem.
  • The forecasts imply that blob-based scaling is the dominant near-term lever for Ethereum throughput.

Load-bearing premise

The rate of improvement observed from recent upgrades will persist in future projections without bottlenecks or changes in network usage patterns.

What would settle it

If TPS growth or fee reductions on L2s or mainnet slow or stop before October 2026 for fees and March 2029 for L2 TPS, the specific convergence dates would not hold.

Figures

Figures reproduced from arXiv: 2606.22206 by Bruce Mizrach, Meghan Ambrosia.

Figure 9
Figure 9. Figure 9: Comparison of Solana Transactions Note: The brown bars denote the total non-vote monthly transaction counts on Solana. The pink bars denote the total vote monthly transaction counts on Solana. Daily transaction counts are obtained from our public Dune query: https://dune.com/queries/6946173 and aggregated monthly. Vote transactions make up an average of 71.5% of total transactions in our sample. Solana’s r… view at source ↗
Figure 12
Figure 12. Figure 12: plots daily TPS for optimistic and ZK-rollups from January 2024 through March 2026. Optimistic rollups grew steadily over the period, rising from 29.67 TPS in January 2024 to 195.22 TPS for March 2026. ZK-rollups followed a different trajectory, reaching 33.57 TPS in March 2024 before declining sharply to around 7.82 TPS in March 2026 [PITH_FULL_IMAGE:figures/full_fig_p019_12.png] view at source ↗
Figure 15
Figure 15. Figure 15: visualizes the TPS of all networks monthly from January 2024 through March 2026 [PITH_FULL_IMAGE:figures/full_fig_p023_15.png] view at source ↗
Figure 19
Figure 19. Figure 19: Ethereum Mainnet TPS: Forecast Using Gas Limit Time Trend [PITH_FULL_IMAGE:figures/full_fig_p029_19.png] view at source ↗
Figure 22
Figure 22. Figure 22: displays the forecasted L2 total TPS under our defined blob count increase structure [PITH_FULL_IMAGE:figures/full_fig_p033_22.png] view at source ↗
Figure 23
Figure 23. Figure 23: presents historical and forecasted TPS for L2 Total and Solana until 2030 [PITH_FULL_IMAGE:figures/full_fig_p034_23.png] view at source ↗
Figure 24
Figure 24. Figure 24: Ethereum Median Transaction Fee and Block Gas Limit [PITH_FULL_IMAGE:figures/full_fig_p035_24.png] view at source ↗
read the original abstract

We study the evolution of transaction speed and fees from January 2024 through March 2026, comparing Ethereum Mainnet and its Layer 2 (L2) networks, as well as Solana and Polygon. Ethereum has undergone upgrades that have increased block size and blob count. These upgrades have doubled transactions per second (TPS) on both the Mainnet and the L2 networks. Mainnet median fees have fallen from over \$2 to under \$0.02, and L2 median fees have fallen more than 95% from \$0.05 to \$0.0015. We forecast that Mainnet median fees will converge with Solana in August 2027, but TPS will remain below 100 until 2034. The L1 Strawmap, proposing EIP-7938, a potential exponential increase in the gas limit, brings the Mainnet to only 100 TPS in January 2028. With continued blob expansion, L2s will surpass Solana TPS in March 2029 and have lower median fees by October 2026.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

2 major / 0 minor

Summary. The paper analyzes transaction speeds (TPS) and median fees on Ethereum Mainnet, its Layer 2 networks, Solana, and Polygon from January 2024 through March 2026. It documents that upgrades have doubled TPS on both Mainnet and L2s, with Mainnet median fees falling from over $2 to under $0.02 and L2 median fees dropping more than 95% to $0.0015. Assuming continued blob expansion, it forecasts that L2s will surpass Solana TPS in March 2029 and have lower median fees by October 2026; Mainnet median fees converge with Solana in August 2027 but TPS remains below 100 until 2034. The L1 Strawmap proposing EIP-7938 is projected to bring Mainnet only to 100 TPS in January 2028.

Significance. If the empirical trends and extrapolation assumptions hold, the work supplies concrete data on the effects of recent Ethereum upgrades (blob expansions and related changes) on TPS and fees, enabling direct comparisons across Mainnet, L2s, Solana, and Polygon. The documented doublings in TPS and sharp fee reductions provide measurable evidence of scalability gains. The forecasts, while dependent on untested persistence of trends, offer a quantitative baseline for assessing future ecosystem trajectories.

major comments (2)
  1. [Abstract] Abstract: The dated forecasts (L2 TPS > Solana by March 2029; L2 fees < Solana by October 2026; Mainnet TPS <100 until 2034) rest on linear extrapolation of observed doublings from recent upgrades under the assumption of 'continued blob expansion.' No explicit functional form, parameter values for the extrapolation rate, sensitivity to demand elasticity, or upper bounds on blob count/gas limit growth are supplied, so the crossing dates are not robust to plausible changes in usage patterns or protocol limits.
  2. [Abstract] Abstract: The specific claim that EIP-7938 under the L1 Strawmap yields only 100 TPS for Mainnet in January 2028 provides no derivation linking the proposed exponential gas-limit increase to this TPS figure (e.g., no accounting for average gas per transaction, block interval, or realized utilization).

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for the constructive comments. We agree that the forecasting methodology and the EIP-7938 derivation require explicit documentation to allow readers to assess robustness. We will revise the abstract and add a dedicated methods subsection with the functional form, fitted parameters, sensitivity checks, and step-by-step derivation. Point-by-point responses are below.

read point-by-point responses

  1. Referee: [Abstract] Abstract: The dated forecasts (L2 TPS > Solana by March 2029; L2 fees < Solana by October 2026; Mainnet TPS <100 until 2034) rest on linear extrapolation of observed doublings from recent upgrades under the assumption of 'continued blob expansion.' No explicit functional form, parameter values for the extrapolation rate, sensitivity to demand elasticity, or upper bounds on blob count/gas limit growth are supplied, so the crossing dates are not robust to plausible changes in usage patterns or protocol limits.

    Authors: We accept that the extrapolation must be stated explicitly. The projections fit an exponential model to the observed TPS doubling between January 2024 and March 2026 and assume the same rate continues under continued blob expansion. In revision we will report the exact functional form TPS(t) = TPS_0 * 2^(t / au), the fitted doubling time au derived from the data, the assumed blob-expansion schedule, and a sensitivity table varying demand elasticity and imposing plausible gas-limit ceilings. These additions will appear in both the abstract and a new methods paragraph. revision: yes

  2. Referee: [Abstract] Abstract: The specific claim that EIP-7938 under the L1 Strawmap yields only 100 TPS for Mainnet in January 2028 provides no derivation linking the proposed exponential gas-limit increase to this TPS figure (e.g., no accounting for average gas per transaction, block interval, or realized utilization).

    Authors: The 100 TPS figure is obtained by applying the exponential gas-limit schedule in EIP-7938 to the observed average gas per transaction (~100 k gas), the 12-second block interval, and realized utilization rates (50–70 %). We will insert the explicit formula TPS = (gas_limit(t) * utilization) / (gas_per_tx * block_time) together with the numerical values and the date at which the limit reaches the level that produces 100 TPS. This derivation will be added to the revised manuscript. revision: yes

Circularity Check

0 steps flagged

No circularity: forecasts are explicit trend extrapolations from observed upgrade effects.

full rationale

The paper reports empirical observations (TPS doubling and fee drops from 2024-2026 upgrades) and then states forward projections under the explicit assumption of 'continued blob expansion.' No equations, fitted parameters renamed as predictions, self-citations, or ansatzes are present in the provided text. The dated crossing points (e.g., L2 TPS surpassing Solana in March 2029) rest on an external rate assumption rather than any self-referential definition or reduction to inputs. This is a standard (if debatable) extrapolation exercise with no load-bearing circular steps.

Axiom & Free-Parameter Ledger

1 free parameters · 0 axioms · 0 invented entities

The forecasts rely on fitted trends from historical data, which act as free parameters in the projections. No other axioms or invented entities are identifiable from the abstract.

free parameters (1)
  • extrapolation parameters for TPS and fees
    Used to project future dates but details not in abstract

pith-pipeline@v0.9.1-grok · 5721 in / 1209 out tokens · 44632 ms · 2026-06-26T10:33:04.108795+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

32 extracted references · 10 canonical work pages

  1. [1]

    Economic Inquiry , volume =

    Donmez, Anil and Karaivanov, Alexander , title =. Economic Inquiry , volume =. doi:https://doi.org/10.1111/ecin.13025 , url =. https://onlinelibrary.wiley.com/doi/pdf/10.1111/ecin.13025 , abstract =

    work page doi:10.1111/ecin.13025

  2. [2]

    2024 , month=

    Alexander Karaivanov and Shayan Zarifian , title=. 2024 , month=. doi:None , url=

    2024

  3. [3]

    Impact of EIP-4844 on Ethereum: Consensus Security, Ethereum Usage, Rollup Transaction Dynamics, and Blob Gas Fee Markets , year=

    Park, Seongwan and Mun, Bosul and Lee, Seungyun and Jeong, Woojin and Lee, Jaewook and Eom, Hyeonsang and Jang, Huisu , journal=. Impact of EIP-4844 on Ethereum: Consensus Security, Ethereum Usage, Rollup Transaction Dynamics, and Blob Gas Fee Markets , year=

  4. [4]

    2020 , month = oct, howpublished =

    Buterin, Vitalik , title =. 2020 , month = oct, howpublished =

    2020

  5. [5]

    2026 , howpublished =

    L1. 2026 , howpublished =

    2026

  6. [6]

    Transaction Speed, Network Activity, and Power Consumption in Ethereum, Ethereum Layer 2, and EVM-Compatible Protocols: A Quantitative Study , year=

    Ginting, Christopher Adiguna , booktitle=. Transaction Speed, Network Activity, and Power Consumption in Ethereum, Ethereum Layer 2, and EVM-Compatible Protocols: A Quantitative Study , year=. doi:10.1109/ICoABCD63526.2024.10704441 , keywords=

    work page doi:10.1109/icoabcd63526.2024.10704441 2024

  7. [7]

    International Journal of Network Management , volume =

    Azevedo Sousa, José Eduardo and Oliveira, Vinícius and Valadares, Júlia and Dias Gonçalves, Glauber and Moraes Villela, Saulo and Soares Bernardino, Heder and Borges Vieira, Alex , title =. International Journal of Network Management , volume =. doi:https://doi.org/10.1002/nem.2113 , url =. https://onlinelibrary.wiley.com/doi/pdf/10.1002/nem.2113 , note =

    work page doi:10.1002/nem.2113

  8. [8]

    Proceedings of the 2022 ACM SIGSAC Conference on Computer and Communications Security , pages =

    Liu, Yulin and Lu, Yuxuan and Nayak, Kartik and Zhang, Fan and Zhang, Luyao and Zhao, Yinhong , title =. Proceedings of the 2022 ACM SIGSAC Conference on Computer and Communications Security , pages =. 2022 , isbn =. doi:10.1145/3548606.3559341 , abstract =

    work page doi:10.1145/3548606.3559341 2022

  9. [9]

    and Rajbahadur, Gopi Krishnan and Hassan, Ahmed E

    Pacheco, Michael and Oliva, Gustavo A. and Rajbahadur, Gopi Krishnan and Hassan, Ahmed E. , title =. Empirical Software Engineering , year =. doi:10.1007/s10664-022-10283-7 , pmid =

    work page doi:10.1007/s10664-022-10283-7

  10. [10]

    A stochastic performance model for evaluating ethereum layer-2 rollups , journal =

    Carlos Melo and José Miqueias and Johnnatan Messias and Glauber Gonçalves and Francisco Airton Silva and André Soares and Jean Araujo , keywords =. A stochastic performance model for evaluating ethereum layer-2 rollups , journal =. 2026 , issn =. doi:https://doi.org/10.1016/j.future.2025.108316 , url =

    work page doi:10.1016/j.future.2025.108316 2026

  11. [11]

    Peer-to-Peer Networking and Applications , year =

    Ta, Minh Thanh and Do, Tien Quyet , title =. Peer-to-Peer Networking and Applications , year =. doi:10.1007/s12083-023-01598-3 , issn =

    work page doi:10.1007/s12083-023-01598-3

  12. [12]

    Blockchain transaction fee and Ethereum Merge , journal =

    Archana Jain and Chinmay Jain and Karolina Krystyniak , keywords =. Blockchain transaction fee and Ethereum Merge , journal =. 2023 , issn =. doi:https://doi.org/10.1016/j.frl.2023.104507 , url =

    work page doi:10.1016/j.frl.2023.104507 2023

  13. [13]

    and Khabbazian, M

    Alizadeh, S. and Khabbazian, M. , title =. EPJ Data Science , year =

  14. [14]

    Ethereum vs Solana: A Comparative Study of Blockchain Architecture on Performance, Security, and Ecosystem Development , journal =

    Michael Wijaya and Franscelino Melvyn and Reina Setiawan and Reinert Yosua Rumagit , keywords =. Ethereum vs Solana: A Comparative Study of Blockchain Architecture on Performance, Security, and Ecosystem Development , journal =. 2025 , note =. doi:https://doi.org/10.1016/j.procs.2025.08.273 , url =

    work page doi:10.1016/j.procs.2025.08.273 2025

  15. [15]

    2025 , month =

    Fusaka Testnet Announcement , howpublished =. 2025 , month =

    2025

  16. [16]

    Benchmarking Blockchain Data Availability: A Comparative Empirical Study of Six Production DA Protocols , doi =

    Kumkar, Prasad and Sarode, Pranjal , year =. Benchmarking Blockchain Data Availability: A Comparative Empirical Study of Six Production DA Protocols , doi =

  17. [17]

    2026 , month =

    Ethereum gas and fees: technical overview , howpublished =. 2026 , month =

    2026

  18. [18]

    2024 , month =

    van der Wijden, Marius , title =. 2024 , month =

    2024

  19. [19]

    2025 , month = nov, organization =

    Chodoła, Kamil and Bermudez, Carlos and Adams, Ben , title =. 2025 , month = nov, organization =

    2025

  20. [20]

    and Chaudhary, A

    Caparros, B. and Chaudhary, A. and Klein, O. , title =. Warwick U. Working Paper , url=

  21. [21]

    IET Blockchain , volume =

    Dyade, Antosh Madappa and Dyade, Maheshwari Antosh and Shinde, Sahadev Maruit , title =. IET Blockchain , volume =. doi:https://doi.org/10.1049/blc2.70014 , url =. https://ietresearch.onlinelibrary.wiley.com/doi/pdf/10.1049/blc2.70014 , abstract =

    work page doi:10.1049/blc2.70014

  22. [22]

    2023 , note =

    Dapplion and Kalinin, Mikhail , title =. 2023 , note =

    2023

  23. [23]

    2025 , note =

    Gold, Sophia and Jayanthi, Parithosh and Wahrstätter, Toni and Beekhuizen, Carl and Dietrichs, Ansgar and Feist, Dankrad and Stokes, Alex and Rudolph, Josh and Rebuffo, Giulio and Slivkoff, Storm and Chodo. 2025 , note =

    2025

  24. [24]

    2024 , note =

    Jayanthi, Parithosh and Wahrstätter, Toni and Calder-Mason, Sam and Davis, Andrew and Dietrichs, Ansgar , title =. 2024 , note =

    2024

  25. [25]

    2025 , note =

    Mackey, Mark and Kripalani, Ra\'. 2025 , note =

    2025

  26. [26]

    2019 , note =

    Buterin, Vitalik and Conner, Eric and Dudley, Rick and Slipper, Matthew and Norden, Ian and Bakhta, Abdelhamid , title =. 2019 , note =

    2019

  27. [27]

    2022 , note =

    Buterin, Vitalik and Feist, Dankrad and Loerakker, Diederik and Kadianakis, George and Garnett, Matt and Taiwo, Mofi and Dietrichs, Ansgar , title =. 2022 , note =

    2022

  28. [28]

    2025 , note =

    Feist, Dankrad , title =. 2025 , note =

    2025

  29. [29]

    2018 , note =

    Akhunov, Alexey and Salem, Moody , title =. 2018 , note =

    2018

  30. [30]

    2024 , note =

    Lightclient , title =. 2024 , note =

    2024

  31. [31]

    2020 , note =

    Vlasov, Alex and Olson, Kelly and Stokes, Alex and Sanso, Antonio , title =. 2020 , note =

    2020

  32. [32]

    2024 , note =

    Ryan, Danny and Feist, Dankrad and D'Amato, Francesco and Wang, Hsiao-Wei and Stokes, Alex , title =. 2024 , note =

    2024