[Constitutional] AIP: Remove Cost Cap on Arbitrum Nova

[Arbitrum]

Constitutional

Note: this proposal will be subject to an individual Snapshot vote; if it receives 95% or more FOR votes on Snapshot and reaches an informal quorum of 100 million, it may be combined with several other maintenance-related proposals for a joint Tally vote.

Abstract

This proposal will remove the Amortized Cost Cap, which is expected to result in higher gas fees for users on Arbitrum Nova. This increase is necessary to ensure the long-term sustainability of the network and reduce the financial burden on the Arbitrum Foundation. Given the diminishing relevance of Nova and the evolving Layer 2 landscape, lifting this cap will prevent inefficiencies and unnecessary constraints on network operations.

Motivation

Arbitrum Nova was initially positioned as a cost-efficient alternative to Arbitrum One, offering 30-50x cheaper transaction fees at launch. However, several developments have significantly reduced Nova’s competitive advantage:

• EIP-4844 blobs: The introduction of EIP-4844 significantly reduced the cost difference between Nova and Arbitrum One, from 30-50x to a mere 2x. This difference is negligible given the already low costs on Arbitrum One.
• Orbit adoption: Orbit chains have become a more attractive option for teams seeking gas sponsorship and customization. Their ease of deployment has led many teams to prefer Orbit chains over Nova.

Given the now-minimal cost advantage over Arbitrum One and the rise of Orbit chains, maintaining an amortized cost cap on Nova — which currently prevents batch posters from recovering full L1 costs — is no longer justifiable. Removing the cap will stop the Arbitrum Foundation from subsidizing these subsidies and align cost recovery with actual usage.

Rationale

The continued enforcement of an amortized cost cap on Nova introduces inefficiencies that do not align with its current usage patterns and network activity. This setting limits the percentage of L1 posting costs that batch posters can recover from the chain (denominated in basis points, or bips, where 10000 = 100%). Currently, the cap prevents full cost recovery, forcing the Arbitrum Foundation to subsidize the difference. Removing the cap (i.e., setting it to 0 bips) would:

• Ensure batch posters are able to recover 100% of their L1 posting costs.
• Eliminate recurring deficits currently paid by the Arbitrum Foundation.
• Reflect Nova’s lower strategic priority within the broader Arbitrum roadmap.

Historical data on Arbitrum Nova batch poster deficits can be seen below:

This data highlights the ongoing inefficiencies and justifies the removal of the amortized cost cap.

Key Terms

• Amortized Cost Cap (setAmortizedCostCapBips): A parameter within the ArbOwner system contract that limits the percentage of Layer 1 transaction posting costs that batch posters are allowed to recover from the network. Expressed in basis points (bips), where 10000 = 100%.
• ArbOwner Contract: A privileged precompiled contract that allows the ArbitrumDAO to configure core protocol settings, such as the amortized cost cap.
• EIP-4844: An Ethereum Improvement Proposal introducing blob-carrying transactions. Its implementation has significantly reduced the cost disparity between Arbitrum Nova and Arbitrum One, undermining the original economic rationale for Nova’s fee model.
• Batch poster transactions: Transactions that aggregate multiple operations and submit them as a single on-chain transaction. With the enforced fee cap, these transactions become less cost-effective under current network dynamics, contributing to operational inefficiencies.

Specifications

The DAO will execute a call to the ArbOwner contract on Arbitrum Nova to set the amortizedCostCapBips value to 0, thereby allowing batch posters to recover 100% of their L1 posting costs.

Steps to Implement

More detailed information about the specific implementation will be provided closer to the date of the formal on-chain Tally vote. This proposal will follow these below steps:

1. An AIP outlining the proposed modification to Arbitrum Nova is proposed to the ArbitrumDAO Forum for discussion (this post)
2. A temperature check vote is held on Snapshot
3. A formal AIP and call-to-vote is proposed to Tally for the on-chain vote
4. Should the Tally vote pass, the removal of amortized cost cap will be executed on Arbitrum Nova following the required waiting periods and phases, as outlined in the ArbitrumDAO Constitution for a Constitutional AIP
5. Monitoring and reporting on the impact post-implementation

Timeline

The following timeline outlines proposed milestones from initial discussions to full implementation. Adjustments may be made based on community feedback and DAO governance requirements.

1. Forum discussion: original date of forum post
2. Temperature check: At least 1 week after forum discussion
3. AIP draft finalization: few days after temperature check
4. On-chain Tally vote: At least 2-3 weeks after temperature check
5. Waiting period on L2: this will last 1 week after Tally vote is finalized
6. Execution of the removal (contract call): soon after the waiting period
7. Post-implementation monitoring & reporting: ongoing, the public dashboard will be updated with new data 1 month after the removal of the amortized cost cap

[cp0x]

First, I would like to outline a few pros and cons:

Pros:

1. Reduction of unnecessary and quite significant costs for Arbitrum.
2. No more need to subsidize Nova, as Arbitrum has already significantly reduced fees thanks to recent updates.

Cons:

• Potential loss of Nova chain users. The chain’s TVL is only $700,934 (according to DeFiLlama), which is very low, and transaction fees are already so minimal that most users may not even notice any changes

Considering Nova’s substantial operational costs, it seems clear that ending the subsidies is the right decision

[DonOfDAOs]

I would almost consider that con, a pro. It seems like there is no reason not to reduce the incentive of Nova and consolidate usership on One.

[GozmanGonzalez]

Thanks for a well-structured proposal.

From a governance perspective, this change meets the criteria of a constitutional AIP and follows the proper process. That said, I’d like to flag a few important considerations before it proceeds to Snapshot and Tally:

1. Governance Integrity: Removing the Amortized Cost Cap shifts the financial burden from the Foundation to users. While this is fiscally logical given Nova’s declining use case, it’s a material change in economic design and should be accompanied by transparent communication to affected stakeholders (e.g., batch posters and dApp teams).
2. Legal and Contractual Clarity: Please ensure that this change:

• Is not retroactive
• Does not affect any prior agreements or commitments involving Nova
• Includes a disclaimer that clearly positions this as a forward-looking parameter update

3. Post-Implementation Monitoring: I support the proposal under the condition that clear metrics are published (e.g., cost shifts, batch activity, user drop-off), ideally via the existing public dashboard.

This is a timely and necessary update given EIP-4844 and Orbit’s evolution. I support the change, provided the rollout is done with accountability and transparency.

[paulofonseca]

@Arbitrum

Note: this proposal will be subject to an individual Snapshot vote; if it receives 95% or more FOR votes on Snapshot and reaches an informal quorum of 100 million, it may be combined with several other maintenance-related proposals for a joint Tally vote.

this is not how we usually do things in here, right?

I think it’s unfair to break precedent to how we’ve been considering that an important snapshot vote is passed.

We should maintain the 3% quorum requirement for this offchain vote as well. Otherwise we are undermining the legitimacy of this decision when compared to all others.

[AbdullahUmar]

Below are the opinions of the UADP:

Removing the Amortized Cost Cap on Nova makes sense from a financial and strategic perspective.

It seems like builders may just focus on launching Orbit AnyTrust chains if they desire low txn cost for high activity applications. Dev optionality is a lot better now than it was a couple of years ago, and Nova served its purpose more prominently as a precursor to Orbit. If I recall correctly, Reddit conducted a points reward campaign using Nova in 2023. That would in part align with the higher levels of Nova DEX activity in the graph below. If the Reddit program or alike initiatives were still active, I’d recommend coordinating with the right stakeholders to ensure that Nova-based programs are not adversely affected—but it seems most programs have already subsided.

The lackluster trailing 12-month activity is strong reason to reduce AF-based subsidies. Builders have moved. Users have moved. And the revenue-to-cost ratio keeps worsening:

One pending question is—how many other resources are being used to run and facilitate Nova? If there are places where AF perhaps could reduce resource allocation to Nova, it may be worth considering.

[Euphoria]

The following reflects the views of the Lampros DAO governance team, composed of Chain_L (@Blueweb) and @Euphoria, based on our combined research, analysis, and ideation.

We support the proposal to remove the amortized cost cap on Nova. Given recent ecosystem developments, particularly Ethereum’s EIP-4844 and growing adoption of Orbit chains, it makes sense to align Nova’s operational costs with its current market position.

We particularly align with the points raised by @GozmanGonzalez regarding the necessity of clear stakeholder communication and detailed post-implementation monitoring. We believe it would be beneficial to outline precisely how communication will be structured to reach potentially affected Nova-based projects and batch posters. This could include specific timelines and dedicated channels, ensuring transparency and minimizing confusion during the transition. This will effectively manage the transition, reduce disruption, and offer clarity on potential migration paths to Arbitrum One or Orbit chains.

[EzR3aL]

The Arbitrum DAO already has a lot costs and Nova is basically irrelevant with its current TVL and the competition in the L2 landscape. As outlined by the Foundation, EIP4844 changed this even more making Nova only a burden that is not bringing any new net user or developer.

I would even go one step further and try to slow down everything related to Nova step by step. Making sure that over time it will “die” and funds to upkeep it can be used for other important upgrades to Arbitrum and its ecosystem.
This should be a well organized off boarding plan to make sure user have enough time to bridge their assets and developer can seamlessly move to Arbitrum One or their own Orbit chain.