Das UberChain

AKA Superchain

Background

XGA is building a platform for creating extensible gas auctions. v1 of XGA is Live on the Ethereum main net (as of June 12th, 2024). read more at docs.xga.com

XGA is designed and architected by 20squares.xyz - a collective specializing in operations research and mechanism design in the crypto space. Manifold Finance codevelops, implements and maintains the infrastructure for XGA in a joint partnership. Manifold Finance has operated the only non-censoring/non-filtering MEV Relay since Day 0 of PoS (i.e. The Merge).

XGA, at its core, is a market for blockspace options and block priority. These options provide the buyer with inclusion rights in non-priority blockspace. Various use-cases are enabled by this, such as 'Preconfirmations' (for L1 Ethereum), support for Multidimensional Markets, and more.

Draft Proposal

We want to join the Superchain, we believe that our solution is both unique in approach and especially relevant towards OP.

XGA is built on top of the OP Stack. No changes for v1 have been made to the current OP stack implementation.

• XGA joins the Superchain and contributes planned changes for upstream consideration ^[1]

• XGA is used as the 'suggested' ^[2] market for providing both aggregation of blob transactions and accepting blob transactions from member superchains (which shall be given priority) for inclusion on Ethereum L1 using the XGA Auction. This is conditioned on meeting forecasted/expected cost savings/user benefit metrics.

  • This means a certain number of blocks per day are guaranteed to include OP Rollup blobs at a better price than what the market would provide.

• XGA provides SLA guarantees regarding outages; remuneration is made using our captive insurance mechanism.

  • This means outages are covered. In the event of service disruption, the blobs would just propagate over the public mempool.

• XGA can collaborate with the OP team in designing a pricing endpoint that can be used by rollups to dynamically adjust their pricing parameters where appropriate.

• XGA can additionally provide 'tailored market access' for larger rollups if need be.

Why is this important for the Superchain?

Currently, there is little incentive for builders to include blob transactions as they incur a latency penalty; thus, including them makes them less competitive; see ethresear.ch/t/blobs-reorgs-and-the-role-of-mev-boost/19783

As more and more OP Rollups are deployed, blob contention and congestion will increase without an adequate market clearing mechanism. This

• Superchain provides a way of aggregating capacity needs from different levels of Rollup usage, enabling better network effects and incentives to join Superchain

Current Blobspace Issues

From the research article "Blobs, Reorgs, and the Role of MEV-Boost" the TLDR is:

• Builders have an incentive to not include blobs because of the higher latency they cause. (We independently came to the same conclusion 8 months ago after discussing this with various builders.)
• Non-MEV-Boost users include, on average, more blobs in blocks than MEV-Boost builders. (This makes sense as they are not latency sensitive, i.e. they build blocks locally.).
• MEV-Boost users show a significantly lower probability of being reorged than Non-MEV-Boost users (see section MEV-Boost and Reorgs for details).
• Rsync-Builder and Flashbots have a lower average number of blobs per block than other builders. (Blob inclusion only happens when network activity is extremely low)

Comparing MEV Boost with Vanilla Block Building and Blob Inclusion

Expected Outcomes

Capacity Planning for Member Rollups

• Capacity Planning enables multiple benefits
  • Cost Savings: Lower transaction and operational costs.
  • Scalability: Ability to handle numerous auctions and participants efficiently.

Reducing user costs^[3] and passing that savings directly to the end users must be a direct benefit that is observed. The Superchain can benefit by virtue of aggregation and achieve larger economies of scale.

Note that this increases the computational complexity of simulating and execution of a block, as well as increases the latency of block propagation. This is where XGA will additionally help by virtue of submitting directly to the Relay we operate as opposed to having to be included by a builder.

Blob Contention Scenario

We define blob contention as # of Blobs in the mempool */* 6

This scenario plays out whenever we see a continuous reduction of blobs being included, creating a backlog for inclusion.

https://www.ethernow.xyz/mempool/all?filter=e5a0a771-2336-4981-ab51-916b5c1e0050

More interesting potential use cases

• Support for introspecting upstream originating L2's, e.g. enabling support for differential costs based on blockspace fungibility (non-fungibility).
  E.g. supporting the Plasma DA Layer: https://specs.optimism.io/experimental/plasma.html

  • redstone, fraxtalda, celestia, eigenda, etc. https://dune.com/queries/3517710

• Inclusion Market can help facilitate blob transaction inclusion without paying more ETH potentially through side channel payments, e.g. in OP token or Rollup specific Gas Token potentially.

  • This becomes more interesting if OP launches a DA layer with OP as unit-of-account
    Adjusting based on conditions such as aggregate demand from a specific Rollup (combinatorial).

• BatchSubmitter
  BatchSubmitter's implements the logic by which the rollup sends either calldata or blob-type data. This can be leveraged for the purposes of fine-tuning fee parameters.

Proposal Methodology

XGA V1 is Currently in the Deployment Phase.
XGA V2 is Currently in the Planning Phase.

We propose an initial evaluation phase of a subset of rollups (ideally a mix of large and small, not the entire Superchain registry!) with our existing platform. Iterating on this, as well as getting feedback and real-world metrics to inform a larger cohort for testing version 2. After the second round of testing, the evaluation of the merits and results of the testing would then be used to inform the decision-making process and governance apparatus necessary to make us a 'suggested provider'.

Typical Methodology Milestones

Planning Phase: Define project scope, objectives, and deliverables.

Development Phase: Develop and test smart contracts and the auction platform.

Integration Phase: Integrate the platform with existing systems and conduct user training.

Deployment Phase: Launch the platform and monitor performance.

Evaluation Phase: Assess outcomes and gather feedback for continuous improvement.

---

1. • TRXID Precompile (intra transaction index)
   • New Data Compression Contract (working with Solady author)
   • Enshrined Gas Token ERC Contract Interface (proposal) (maybe LST's too)
   • STATICCALL Support
   • L1BlockSnapshotter (allows for the creation of a snapshot of the L1 blockhashes for historical retrieval on L2)
   ↩︎

2. This is not a per se 'default'; it would be an option to 'opt into' for Superchain rollups. ↩︎

3. This also includes the rollup operators themselves, as it offsets rollup costs relative to the Superchain revenue share that they would pay out. ↩︎