Ecosystem Brief: Rollup Specialization

Key Trends

Rollup Specialization

General purpose chains are designed to accommodate the widest set of smart contract applications, including financial applications. Financial use cases require significant security to back the assets, and this requirement results, in Ethereum’s case, with relatively high transaction costs. These high transaction costs, while to the benefit of financial protocols, is a detriment to consumer protocols, for example, which consist of high volume, low-value transactions. Even Ethereum's rollup transaction cost of a few cents is prohibitive to many high-volume, low-monetary-value use cases such as social and gaming.

In order to overcome the restrictions of general-purpose chains and rollups, it has been a long-held theory that application protocols would eventually split off from their smart contract roots and operate their own app chain in order to optimize the blockchain architecture for their particular use case. This thesis manifested with dYdX’s migration from an Ethereum rollup to its own Cosmos-based app chain so that it could internalize its order book matching engine within the validator design.

However, there are tradeoffs when migrating to an isolated chain of your own, namely the loss of liquidity network effects, cross-protocol composability, and potentially developer tooling. Therefore, it is only a viable strategy for large, mature applications that have existing brands, user relationships, and product-market fit.

In contrast to the either-or option of general purpose protocol or app chain, a third design has emerged as a trend over the last few months – sector-specific chains.

Sector-specific chains are optimized for the core requirements of a particular family of applications instead of a single application. Games, for example, are likely all share similar requirements, such as high-volume, low-value transactions. By building optimizations targeted at a particular application vertical, the success of the chain is not dependent on the success of a singular application and is also able to share the infrastructure cost across multiple applications, benefiting all. This meets the polar options of a fully generalizable chain or a dedicated app chain in the middle as applications get both the base optimizations needed for their functional area as well as the benefits of shared infrastructure.