Discover how Celestia’s modular approach to blockchain helps solve the scalability problem through data availability sampling.
Share
Subscribe to the AlphaWire Newsletter
Blockchain networks are built to work en masse — a decentralized, global solution to many of the world’s financial, supply, and governance problems. But while that vision is sound, the technology still has a way to go regarding scalability.
Most blockchains follow a “monolithic” architecture, one where transaction settlement, execution, and data availability all happen on one layer. All of that is a lot to handle, leading to network congestion, which limits scalability. Ironically, the more popular a network gets, the worse it may perform, unless the correct infrastructure is in place.
Many solutions attempt to address blockchain limitations, like scalability solutions such as sharding and rollups, or alternative consensus mechanisms like proof-of-stake (PoS). However, many of these solutions introduce the data availability problem.
Say you want to run a full Bitcoin node (become a block producer). You need to download the whole blockchain’s history, as you need all of its data available to you. When a node broadcasts a transaction, you and every other node check that transaction against Bitcoin’s history, hence the network’s longer transaction validation times.
You can also run a light node, which can send and receive Bitcoin without downloading its entire history. Instead, light nodes download block headers, essentially summaries of full blocks. They then rely on full nodes to receive block data.
This works on a monolithic network like Bitcoin, which doesn’t require instantaneous transactions. However, downloading a network’s full transaction history becomes problematic when it comes to scalability: the process is computationally expensive and time-consuming as a network grows, limiting node participation to those with the space and bandwidth for it.
Many scalability solutions, like sharding or layer-2 rollups, rely on partial node validation. Processes where nodes don’t need a network’s entire history to function. But how do they ensure security?
Also, the data availability problem goes beyond nodes simply downloading a blockchain’s entire transaction history. It’s about nodes being able to publish and access all of the transaction information they claim to have. If full nodes don’t have all of a network’s data, malicious block producers can lie about transactions they’ve published, destroying the trustless nature of blockchain and leading to issues like the double-spend problem.
It’s like if a bank could move your money around without showing you proof. Sure, the bank can say it moved money to pay for one of your bills, but if you don’t see the transaction record, you can’t know for sure. The bank could claim it processed your transaction while using the funds for something different.
If block producers aren’t required to publish all transaction data, the same problem can occur. But requiring block producers to publish all of that data takes time and computational power, as does ensuring that every other node can validate it.
So, how does a network scale without sacrificing security? Blockchain network Celesia believes the answer lies in a modular approach.
Pitched as a “data availability network that securely scales with the number of users,” Celestia tackles the data availability problem by focusing on it alone. It separates transaction settlement and execution into their own chains, emphasizing a process called data availability sampling (DAS) instead of trying to be an all-in-one solution.

This modular approach breaks up the following core blockchain functions:
By ensuring each process is on its own chain, Celestia provides optimized versions of each function for other networks. However, it’s important to note that modular blockchains are more of a mix-and-match toolkit. Developers can use a consensus and execution layer without a settlement layer if they prefer.
Alongside breaking up each blockchain function, any node from a light node to a full node can process transactions on Celestia. However, the network’s main draw is its ability to function primarily via light nodes, which is where data availability sampling comes in.
For DAS to work, light nodes sample random bits of block data from across the network rather than downloading whole blocks. As light nodes continue to complete their data sampling, their confidence in available data will increase. Once a light node hits a certain confidence level, it will broadcast that the data is available.
In theory, the network would have more than enough nodes sampling small amounts of data to ensure none of it is being withheld by a bad actor. In fact, as more nodes join the network, more sampling checks come to be, meaning the network will increase in security as it scales.
Essentially, Celestia creates a system where verification is a shared, scalable process rather than requiring every node to have the full ledger. From there, it presents transactions in the order they were made, while ensuring all data is always available.
The process works through a combination of nodes:
With light nodes making up most of the data availability layer, what about the others?
While Celestia will primarily focus on data availability, part of the project’s pitch is its integration with other networks.
For example, a decentralized app (dApp) developer may choose Celestia’s consensus and data availability layers while taking advantage of Optimism’s rollups for execution and Ethereum for settlement.
The idea is for developers to choose their top functions, harnessing what they need to best deliver their services. A simplified, modular dApp development method. The network isn’t trying to be everything. It’s trying to be the best at one thing, while supporting the best of the rest.
This modular approach, in theory, brings countless benefits to dApp developers:
Celestia already supports dozens of apps, and with just under 400k followers on X, the network has undoubtedly made a name for itself. Its presence as a modular blockchain seems to be growing, and so is its number of integrations.

Of course, no solution is without its challenges. Celestia’s method solves some problems while introducing others:

Celestia resembles Amazon Web Services in a few key ways, largely by providing infrastructure so developers can focus on apps. Before Amazon’s cloud computing initiative, web-focused companies had to build their own servers. Server hosting is expensive, especially at scale, so Amazon built cloud computing as a service. While blockchains don’t require servers, they do require consensus and security. Building a decentralized foundation takes time, money, and expertise.
Both AWS and Celestia lower the barrier to entry for app developers, and the latter hopes to deliver where the former revolutionized. Celestia’s modular approach is certainly unique and may popularize a new form of development. However, AWS took off because of its simplicity. The question remains: can Celestia achieve that same level of accessibility?
If it can, Celestia won’t just be a part of blockchain’s future. The network will help define it.
Share
