Sei has built-in features such as parallel order execution, an on-chain order book, and a native order-matching engine that will enable a new suite of financial products. Sector-specific chains optimize their stack for a whole vertical, which is ideal for building a community and functioning as the liquidity hub.

Web3 activity might be showing support for the “fat app theory” as apps on a protocol end up in competition with them. Application-specific blockchains are optimized for a single application as opposed to general-purpose blockchains, which try to provide for multiple uses. Sei, a DeFi-focused Cosmos chain, calls itself a sector-specific chain and lies in the middle of this spectrum. These chains target a sector of apps, such as DeFi or gaming, combining specialized features with general-purpose functionality.

App Chains and Dedicated Blockspace App chains might be positioned to have better economies than general-purpose chains. The "fat protocol thesis" posits that most value generated by application-level activity on a blockchain should flow to the underlying protocol. This thesis emerged from observations of infrastructure and application cycles in the Web2 era. However, the dynamics of the Web3 ecosystem do not entirely align with the fat protocol thesis.

Certain scenarios, such as the competition for blockspace and the resulting negative impacts of maximal extractable value (MEV), create a competitive environment instead of a symbiotic one.

These scenarios support the idea of a "fat app" thesis instead, i.e., the app chain thesis. App chains are independent blockchains that typically have shared security, provided by shared validator sets. Cosmos chains, Avalanche parachains, or Polkadot subnets can be considered app-specific chains. Because these applications have their own chains, they don’t compete with each other for blockspace. Due to their modular software stack, they are free to upgrade and set technical parameters and fee structures, unlike applications that are "renting" space on a general-purpose base layer. dYdX started off on StarkNet, an Ethereum L2, with the vision of building a scalable, decentralized off-chain order book.

But when the dYdX team ran into Ethereum’s throughput limitations, they decided to pivot to an app chain. dYdX on an L2 could only handle 10 trades per second, which is a significant limitation when considering that exchanges like Nasdaq can handle more than 1,000 trades per second. As an independent app chain on Cosmos, dYdX may want developers building on it. However, any DeFi app built on the dYdX chain would be in direct competition with dYdX, which could limit innovation on the chain and increase competition from other app chains. Sei: Going Sector-Specific The give-and-take between general-purpose and application-specific blockchains led to the development of sector-specific chains.

This brings us to sector-specific chains. Because of their position as a “more general but focused” chain, sector-specific chains serve as the liquidity hub for DeFi apps on them. Because app chains only support a single app, bootstrapping a community takes time. On the other hand, general-purpose chains build a community from multiple sources, and users from various platforms can then cross-pollinate. Sector-specific chains fall in the middle. Right now, Sei has 100+ projects in queue to build on it once live. Because of the optimizations Sei has implemented, new DeFi products are possible. Key Architecture Decisions Sei is built with the Cosmos SDK. Leveraging the SDK, the chain uses customized Tendermint for consensus while maintaining IBC compatibility. Sei has many features that make it competitive with existing DEXs, including: Parallel order execution Apps can process orders in parallel “optimistically,” which is later accepted or rejected by consensus. Additionally, block proposers will only send minimal data to include in a block. Both of these alterations help reduce the time to finality, the time between when blocks are committed to the chain. Built-in on-chain order book Built-in order book logic allows for scale unattained by DEXs today. Specifically, Sei uses a central limit order book (orders maintained by a central authority), which is used as default in TradFi.

This system optimizes order placement and matching. On-chain order books have been a challenge because distributed systems tend to be less efficient and will eventually drive up costs at scale. For example, Serum had introduced an on-chain order book as well, but it is on the Solana network instead of a specialized chain. Sei, however, has configured consensus optimized to keep trade costs low and efficient. Because central limit order books are maintained by a central authority, they’re not quite as aligned with DeFi principles as automated market makers (AMMs) such as Uniswap. While the former is much more efficient, apps on Sei might see a struggle in adoption because of this tradeoff. On the other hand, centralization can be attractive to entities trying to remain compliant. Native order matching engine Trades will be filled with Frequent Batch Auctions, helping with price fairness and frontrunning prevention. To avoid “bad” MEV (i.e. frontrunning or sandwich attacks), Sei aggregates market orders and executes them at the same time and price. Fast finality Because of the design decisions Sei made, it achieves finality in 600ms.

Injective, another Cosmos DeFi chain that can be classified as sector-specific, achieves 1s finality. Both of these sector-specific chains are some of the fastest among any blockchain. Sector-specific blockchains can optimize their infrastructure based on the types of transactions they will process. Sector focus provides the best infrastructure for applications based on the types of transactions they will process. For example, Saga, an out-of-the-box app chain provider, has chosen to begin working with gaming and entertainment teams first. Because gaming applications see high-volume and low-value transactions, the best setup for Saga would be to ensure teams have dedicated and cost-efficient blockspace. Parting Thoughts Sector-specific chains provide apps with desirable configurations to host themselves and their communities. Because of the broadened user base, there is an increase in liquidity in the system. Sei had plans to go live at the end of 2022, but the network is still on testnet. Considering the number of launch partners Sei has secured, it will likely be a strong contender for competing DeFi chains on Cosmos and existing applications outside the ecosystem.