Why rollups-as-a-service will be to Web3 what AWS was to Web2
Amrit Kumar
7 min read
When Amazon Web Services (AWS) burst into the scene in 2006, it quickly became apparent that businesses conducting every element of their operations in-house in the name of self-sufficiency weren’t necessarily working efficiently or smartly. Being able to remotely host applications, services and data became extremely powerful.
Prior to AWS, you had to hire a team of information technology professionals, rent physical space, and invest in hardware to be able to manage your computing and storage needs. The entire set up was completely inelastic. With AWS, upfront costs for compute or storage needs became a thing of the past. By accelerating time-to-market, providing global reach, speed and flexibility, AWS gave their customers a competitive edge, while lowering the barriers to entry for fledgling startups.
Fast forward to now, rollups-as-a-service (RaaS) is going to give blockchain-based applications a similar competitive edge. Rollups add an elastic execution layer to efficiently boost scalability and performance as and when needed — and now it can be done without the higher costs, the chain limitations or the build-time.
In short, RaaS will likely be to Web3, what AWS was to the Web2 era.
An ‘AWS moment’ for the blockchain world?
As it stands, the blockchain ecosystem is plagued by many of the same issues as the early internet and the entire space would greatly benefit from an “AWS moment.” Scalability remains a prime challenge. Developers used layer 1s such as Ethereum, Solana and Avalanche to deploy their apps but found none customizable enough. For instance, a gaming studio building a fully-on-chain game would like the ability to deploy a smart contract larger than the limits imposed by the layer 1. Also, they would prefer to trade off security for latency. For example, Ethereum’s current block time averages about 12 seconds, which may not be sufficient for most on-chain games. Due to this, certain application developers moved away from an existing chain and built their own application-dedicated solutions. CryptoKitties and Axie Infinity were among these.
One can thus certainly draw parallels between the “building your own chain” phenomenon and the pre-AWS era. At that point, projects like Cosmos SDK and Polkadot Substrate tried to fill the gap with SDKs, which reduced the time to launch from several years to a few months — thus creating a massive improvement in the space.
After all, it seems like a no-brainer that one could enhance speed, delivery and quality by creating a chain tailored to the needs of a particular application or product. Undoubtedly, it also makes the developer in question more autonomous — giving them full control of the underlying tech and as an extension, enabling faster implementation of upgrades and security changes. If positioned right, it could also offer dApps a competitive edge in the market, attracting value to both its core layer and to the ecosystem it supports.
A more recent example would be the decision by dYdX — a major decentralized derivatives exchange on Ethereum — to move to its own chain built using Cosmos SDK. The announcement was made in June 2022. While the project is still in full swing, the most recent development update estimates that the mainnet will be ready sometime in September 2023 — i.e., it will take over a year to build an application-dedicated chain using state-of-the-art tooling. It seems like we are back to square one. These SDKs, while extremely helpful, are still not an ideal offering.
Clearly, these projects came with their own set of challenges. They were often too low-level for developers wanting to launch a chain. Another issue was bootstrapping the network, and creating a viable enough ecosystem for that to be successful.
It seems we don’t have an AWS for blockchains yet. One could even go so far to say that each of these chains is building its own AWS, creating more chaos and utilizing massive resources in the process.
How rollup-as-a-service works
Rollup-as-a-service (RaaS) — a type of layer-2 service that helps users to easily create and manage rollups based on their specific needs — is emerging as a viable technology. Modeled as an elastic, pay-as-you-go service, RaaS is enabling faster and cheaper transactions with little to no congestion, by moving some of the computational load off the main network.
Inspired by software-as-service, a RaaS product offering allows developers to launch a rollup by providing a simple and easy-to-use graphical interface. Let’s take the example of a developer wanting to spin up a rollup tied to Ethereum. Instead of using complicated SDKs, one would instead be offered an interface to allow chain customization via simple buttons and clicks. Overall, the time to go live will be reduced from several years or several months to a few minutes.
The other benefit of rollups over sovereign chains is that rollups derive security from an underlying layer 1 and as a result, these rollups are secure even when they only consist of a single node processing user transactions. This is because, in scenarios where the node behaves maliciously, the underlying layer 1 can detect and penalize the node, creating a financial deterrence to misbehavior. As a result, the team launching a rollup does not necessarily need to convince hundreds of validators to oversee and maintain the network. And this reduces the effort needed to build out the ecosystem to run the rollup.
The fact that these rollups are secured by the underlying chain as opposed to sovereign chains, these rollups also offer a unique and powerful feature that sovereign chains can’t: the ability to dispose of the rollup.
A developer expecting an increase in demand for his application could: (1) spin up a fast and scalable rollup secured by a layer 1; (2) use the rollup for as long as needed, and then (3) dispose of the rollup by doing an “end-of-life” settlement on the layer 1, whereby all the assets on the rollup such as NFTs and tokens move to layer 1. This is not possible to achieve with a sovereign chain as by definition, there is no base chain to move the assets to. In fact, these kinds of disposable rollups make the entire system highly resource-optimized. The rollup and its resources are called upon only when the decentralized app expects a considerable demand that a layer 1 can’t handle and once the demand tapers off, the dApp can move back to the layer 1.
In the future, RaaS will likely be to Web3 what SaaS-based models like AWS were to Web2. In fact, Ethereum is one of the leading blockchains prioritizing the development of its layer 2 space to allow for greater scalability and functionality of dApps. Several projects are building rollup-as-a-service solutions but are still in very early stages of development. Some are also building intuitive and easy-to-use dashboards that can help anyone (not just developers) deploy rollups in around 10 minutes. This is lightning-fast compared to options like sidechains and SDKs — and its value proposition is thus immense. Not only could businesses get their required surge in scalability and security, but they could also do so independently.
Of course, one must be prepared for the fact that rollups are in an early stage of development. They might also be limited to the ecosystem they choose to build on, and creating interoperability with other blockchains might be a longer-term goal for some. But as it matures, RaaS can be a catalyst for companies in need of tailored scaling, doing what SaaS did for Web2.
Roll-ups-as-a-service will transform the way enterprises meet scalability demands, inspiring the next era of business transformation. Additionally, the customizable aspect will better satisfy the requirements of consumers that exist in niche pockets. By celebrating what a decentralized design could do for everyone, let us share in the responsibility to shape an internet that is at once personal and connected.
