The Appchain Galaxy: Where Everyone Gets a Blockchain
Technologies and Implications of Appchains
Web3.com Ventures Original Research Analysis
Ethereum is great — for most people at least. But one of the key problems of Ethereum and other public-facing blockchains is that there’s not that much customizability. Sure, you can write and deploy smart contracts, but as a single user or company, you can’t really do much about the consensus mechanisms, execution engines, and other underlying architecture. After all, you aren’t the only one using Ethereum, so you don’t really get to dictate the rules.
Suppose that you are building a cross-chain decentralized exchange that needs to aggregate data from all these different chains , or you want an on-chain real-time-strategy game , or you’re a Fortune 500 company tracking your supply chain. Ready-made public blockchains can’t really handle the thoroughput, data aggregation, or privacy requirements that these applications need. This is where the concept of an “appchain” comes in. Appchains are decentralized, application-specific blockchain networks that are designed to support specific use cases and meet specific requirements. Within this essay, I will explore three exemplary approaches to the emerging appchain landscape: the Cosmos ecosystem, Hyperledger Fabric, and Rollups-as-a-Service (RaaS), before discussing appchains’ rise affects the Web3 infrastructure landscape at-large.
The Cosmos network is perhaps one of the most important appchain solutions within Web3. Stretching all the way back to 2016, the Cosmos network aims to create an “internet of blockchains,” or the infrastructure of a multi-chain world. Essentially, Cosmos provides an software development kit (SDK), which provides developers with the tools and interfaces they need to easily build their own blockchain applications. While the SDK has some default architectures that come out of the box, such as the Tendermint PoS consensus protocol, developers are given a large leeway to overwrite these default architectures and design a blockchain complete with its own custom consensus mechanism, token standards, and smart contract platforms.
Within the Cosmos Network, there are two distinct types of blockchains: “hubs” and “zones” . “Zones” are fairly straightforward to understand — these are just the application-specific chains that run the application (such as a decentralized exchange or on-chain game) itself. “Hubs” on the other hand, function to connect these “zones” and other “hubs” together in order to enable inter-chain communication through the “Inter-Blockchain Communication Protocol” (IBC).
At the center of the Cosmos network is the “Cosmos Hub,” which is the first “Hub” chain to be built on the Cosmos network, and essentially acts as a “Grand Central Station” for all other hubs and zones. Nonetheless, as shown in the diagram above, “Hubs” don’t necessarily need to connect to the Cosmos Hub. A private company, for example, could choose to launch its own private hub that connects its own private zones.
Indeed, the Cosmos SDK has a focus on customizability and interoperability, and developers are given a wide latitude to let their imagination run free. Within the Cosmos ecosystem are some major general-purpose L1 chains, such as the EVM-Compatible BNB Chain (originally Binance Smart Chain) . For an appchain builder, the interoperability of the Cosmos network is incredibly enabling, as it allows the appchain to access data from a large variety of chains through the IBC. This is one of the reasons why many appchains built using the Cosmos network are oracles, multi-chain DEXs, and other services that require the aggregation of data from a wide variety of different chains.
Another way to implement appchains is through Hyperledger Fabric, an open-source blockchain platform for building enterprise-grade applications and solutions. It is one of the projects under the Hyperledger umbrella, which is hosted by the Linux Foundation .
Like the Cosmos SDK, Hyperledger Fabric is designed to support modular blockchain network architecture and provide flexibility in building and deploying blockchain-based applications. It provides a plug-and-play architecture that allows components such as consensus, membership services, and smart contract execution (chaincode) to be easily added or replaced. But the most important difference between the Cosmos SDK and Hyperledger Fabric is one of audience: Cosmos SDK is oriented towards Web3-native users and developers, whereas Hyperledger Fabric targets enterprise-grade institutional clients. Thus, while Cosmos SDK focuses on building a decentralized, interoperable system, Hyperledger Fabric focuses on building private and permissioned networks, where participants must be explicitly authorized to join and participate in the network.
Indeed, enterprise-grade appchains operate in a vastly different way to Web3-native appchains. Whereas Web3-native appchains often focus on the tokenomics and use the blockchain structure to financialize their products, enterprise-grade appchains focus on using the blockchain as an efficient way of record and maintain data. Fundamentally, these enterprise-grade appchains leverage the fact that blockchains are tamper-proof, verifiable, append-only data structures that can have data-writers spread out across time and geographic locations. Essentially, a Hyperledger Fabric-based appchain simply seeks to replace the dusty, inefficient databases enterprises have maintained for decades.
As shown in the diagram above, essentially all of the “permissioned nodes” on the Hyperledger Fabric blockchain are employees in various parts of the supply chain, using the blockchain as a way of synchronously recording different data states — such as if a piece of produce has entered a factory or a store. There is far less of a focus of interconnectivity between different blockchains, and unlike Cosmos, there is no native support for inter-blockchain connectivity using Hyperledger Fabric. Instead, there are data APIs and interfaces for integrating with other blockchain networks or systems. This means that it is possible to build inter-blockchain communication solutions using Hyperledger Fabric, but it requires custom development and integration, unlike the built-in hub-and-zone design of the Cosmos ecosystem. Intuitively, this makes sense: as an enterprise-grade appchain, blockchains built using Hyperledger fabric are usually siloed away from the outer world — after all, you don’t want the public to be messing with your supply chain data.
One case study of Hyperledger Fabric in action is in managing Walmart’s supply chain. Walmart’s appchain, also known as the “Walmart Food Traceability Platform,” is a blockchain-based food traceability system designed to improve food safety and transparency in the supply chain . The Walmart appchain allows for the tracking and tracing of food products from farm to store shelves, helping to quickly identify and isolate any potential food safety issues. At every stage in the process, suppliers had to upload labels and certificates of authenticity through a web interface onto the blockchain, where it would be permanently accessible to all of the relevant stakeholders. With the blockchain serving as a single “ground truth” of information, this greatly reduced the time to record and track food quality issues, such as food-borne diseases:
The Hyperledger Fabric blockchain-based food traceability system built for the two products worked. For pork in China, it allowed uploading certificates of authenticity to the blockchain, bringing more trust to a system where that used to be a serious issue. And for mangoes in the US, the time needed to trace their provenance went from 7 days to… 2.2 seconds! — Hyperledger Foundation 
Thus, Hyperledger showcases the use of appchains in an entirely different setting to Cosmos — to large, enterprise-grade use-cases where the blockchain is used as an alternate to traditional data-recording and tracing methods. Sure, one can argue that this use-case of appchains is “not really Web3,” simply because it doesn’t leverage tokenomics, decentralization, and crypto-native principles in the same way. But the enterprise adoption of appchains cannot be ignored — the underlying technology is the same, and it is a vital step to gaining public trust of blockchain technologies to enable mass adoption. And in the long run, if companies like Walmart have an appchain setup, it is far easier to integrate them into the “Web3-native” world.
A more recent alternate approach to building appchains is through the use of rollups, especially through “rollups-as-a-service” (RaaS). Rollups-as-a-service are a type of blockchain scaling solution that allow for off-chain computation and storage while maintaining the security and trust of the underlying blockchain. They work by batching a large number of transactions into a single, compressed transaction that is then recorded on an underlying L1 blockchain, most commonly Ethereum.
Rollups can be used for appchains because they provide a way to scale the processing and storage of transactions for dApps built on these chains, without sacrificing the security and trust of the underlying blockchain. While the developer may sacrifice some customizability in the underlying consensus layer compared with a Cosmos-based solution, there are several advantages of using a rollup-as-a-service for an application-specific blockchain instead of a full chain.
Perhaps the most important advantage is that there is no need for developers to bootstrap a validator set. Because the consensus and settlement is outsourced to an underlying L1, this results in not only a cheaper appchain solution, but also one that is more robust, as it is secured through the underlying L1. Moreover, because rollups are vertical solutions that are modular in structure, they are not constrained to a single blockchain framework, and can leverage data and functionality on multiple blockchains.
Essentially, in the rollup appchain model, L1 chains become the equivalent of “Hubs” that contain wealths of data and security. A gaming appchain, for example, could use a rollup-based solution where Ethereum’s security is leveraged for consensus and settlement, and Solana’s high-thoroughput can be leveraged for executional speed .
Rollups-as-a-service companies provide the infrastructure and services necessary for appchain developers to implement their custom appchain rollup solution. Different RaaS abstract this toolkit to different levels: some projects, such as the Op Stack , are SDK based, allowing the developer to customize the rollup fairly extensively, other projects such as Constellation focus on a complete no-code “white-glove” deployment solution, allowing appchain developers to focus on the “app” part of development, rather than the “chain” part of development .
The Implications of Appchains
The emergence and rise of these appchain solutions has several profound implications for the Web3 infrastructure landscape at large. Appchains usher in a paradigmatic change between the relationship between “chains” and “apps” from a supply-driven model to a demand-driven model.
Before, when infrastructure solutions were still maturing, and many modern infrastructure concepts such as Proof of Stake, sharding, and rollups were still in the proof of concept stage, infrastructure projects were often implemented for the sake of exploring the feasibility of these technical solutions. This is most typically the case in “third-generation blockchains” such as Avalanche, Cardano, NEAR, and Solana. The supply of these chains (their thoroughput and technical implementations) drove the demand for these chains. In other words, the design of the “chains” determined the the design of the “apps.”
However, the rise of appchains marks a distinct fourth-generation of blockchains, where demand for blockchains drives the supply of these blockchains. All of the technical routes presented above, the Cosmos network, Hyperledger Fabric, and Rollups-as-a-Service, present plug-and-play blockchain deployment solutions that are simple, customizable, and cost-effective. It has never been easier to design, customize, and deploy your own blockchain. Because of this, it is now the opposite way round: the design of the “apps” determines the design of the “chains.”
One likely result of this is that there will likely be a more concentrated L1 scene, with a shift from a monopolistic competition-style scene (where different L1s are slightly differentiated between one another) to an oligopolistic scene, dominated by a few key chains with enormous ecosystems and liquidity. This is because with the advent of appchains, there is less of a reason for top-level DApps to deploy on smaller L1s rather than run their own chain (with infinitely customizable features) or deploy an appchain rollup on a large, established L1. Smaller L1 chains simply cannot compete with the combination of large L1s and appchains in terms of flexibility, customization, security, and support.
Furthermore, appchains are also a key mechanism to institutional onboarding to Web3. As mentioned before, though enterprise chains such as Hyperledger are usually excluded from our typical idea of Web3, they leverage the same underlying blockchain technologies to achieve greater transparency in logistics and supply management. So why can’t this also be extended into payment and beyond? Thus, I believe that it will only be a matter of time before these enterprise-grade blockchains become more integrated with native Web3 DeFi, payment, and NFT projects. Already, there is some exploration into the feasbility of expanding the Cosmos IBC beyond just Cosmos appchains, connecting platforms such as Hyperledger Fabric and targeting other enterprise-grade blockchain solutions .
Nowadays, it is clear that there never was, and never will be a one-size-fits-all blockchain. Whether it be through Cosmos SDK, Hyperledger Fabric, or Rollups-as-a-Service, the maturity and ease of deployment for custom appchains will transform the Web3 infrastructure space into an app-first, user-first experience — a sign of growing maturity in Web3. The future of Web3 space is a multichain galaxy, and given their customizability, effectiveness, and ease of deployment, appchains will play an ever-increasingly important role in this universe, shining like stars in a galaxy of decentralization.
📅 6 February 2023
 Osmosis, a cross-chain DEX: https://osmosis.zone/
 Curio.gg, on-chain gaming company: https://www.curio.gg/
 See BNB Chain Whitepaper: https://github.com/bnb-chain/whitepaper/blob/master/WHITEPAPER.md
 Walmart case study: https://www.hyperledger.org/learn/publications/walmart-case-study
 Example from https://messari.vercel.app/article/the-rollups-as-a-service-ecosystem
 See https://www.optimism.io/op-stack
 See https://constellation-labs.gitbook.io/constellation-labs-documentation/getting-started/overview
Disclaimer: the information presented above is purely educational does not constitute financial advice, and represent the views of the author alone.
Many thanks to my friends at Constellation Labs for their insightful conversations on Rollups-as-a-Service.