[Intern Research] The Hidden Costs of Modular Systems (Summarized)

Summarizing Multicoin's latest research paper on modular chains: Key insights and takeaways for a comprehensive understanding of the topic

Summarizing Multicoin's latest research paper on modular chains: Key insights and takeaways for a comprehensive understanding of the topic.

Modularity Comes with Hidden Cost

• Dev complexity

  • Modular systems such as OP stacks add social and technical complexity to the development of their underlying rollup.

  • Software systems tend to integrate over time and minimize performance and dev complexity, for example, cloud backends, OSes, database engines, and game engines.

• Increased TX cost for users

  • The more asset ledgers (L1 & L2) = more costs for users

  • Moving assets across different rollups via bridges incurs more TX costs

  • Fragmented liquidity = worse trading experience

  • Believes that cross-chain arbitrage is a wasteful use of block space, and most assets should be concentrated in single asset ledgers (single L1 or L2) for maximum efficiency

Limitations of Modular chains

• Modular chains don’t execute code Faster

  • Modular chain = having a separated set of nodes for Data availability (DA) and a separated set of Nodes for execution.

  • Separation of DA and execution doesn’t increase performance, as the underlying hardware (nodes) remain the same.

  • It does reduce the cost of computing by centralizing execution.

• App-specific modular chains allow protocols to capture MEV, but there are easier ways to do it

  • A common belief is that App developers will eventually build App-specific roll-ups to capture MEV back to their tokens.

  • Encode logic in smart contracts on a general-purposed chain achieves the same result but is more efficient.

• Modular chains do not address cross-app congestion

  • A common belief is that app-specific modular chains ensure a given app isn’t impacted by a gas spike by other on-chain activity (such as a popular NFT Mint).

  • App chain doesn’t solve the issue as:

    • L2s pay fees to L1, and when L1 fees spike, L2 fees spike as well (Arbitrum and Optimism gas cost $10 during PePe hype)

  • The only solution to gas spikes is to localize the fee market (separate fee usage among DeFi, NFTs, etc., so that different sectors won’t interfere).

Most apps don’t need modular chains

• Modular chains offer flexibility at the cost of increased complexity.

• Authors argue that there are about 3 specific use cases that require flexibility by modular chain, while most other apps don’t need such flexibility at all.

  • Hot State (dYdX, Sei, dFlow, Pyth)

  • Consensus modification (Osmosis, Thorchain)

  • Threshold signature scheme (TSS) systems (Sommelier, Thorchain, Osmosis, Wormhole, Web3Auth)

• All of the above-mentioned are infrastructure, except dYdX, Osmosis, and ThorChain, which are all DEXes.

• Argues that standalone DEX chains only make sense if they are derivative-focused, such as dYdX and Sei.

Disclaimer

This article is for informational purposes only and should not be considered financial advice. While the author has made a significant effort to ensure the accuracy of the information provided, the author does not explicitly or implicitly guarantee the accuracy or completeness of the information and is not legally responsible for it. The author is also not liable for any losses from using this article for investment or other decision-making purposes.