Every blockchain has to perform 3 core functions; consensus, execution and data availability.
Consensus is how everyone in the network agrees on what transactions have happened and in what order, thus ordering the transactions into a history of events.
Execution is the interpretation of those transactions. Determining what transactions are valid and how they impact and update the state.
A state or state transition is something that occurs each time a block is accepted in the blockchain.
Data availability refers to the ability for transaction data to be made available for nodes to download so that anyone in the blockchain can identify and read the state transition.
Celestia separates these core functions.
The official website describes Celestia as a modular consensus and data network, built to enable anyone to easily deploy their own blockchain with minimal overhead
Now what does that mean?
Celestia is the first modular blockchain network, and modular blockchains are essentially what cloud computing is in Web2.
Every blockchain that has ever been built so far is monolithic or uses monolithic architecture, meaning that the core functions mentioned before, consensus and execution, both happen at the same time and are enforced by the same set of validators. Monolithic architecture is, in a way, a bottleneck that limits scalability as a full node has to execute every single transaction.
Celestia, being a modular blockchain network, uses modular architecture which separates or “decouples” the blockchain stack into more specialized components. Meaning that by using modular architecture, the core functions, consensus and execution, are separated into different layers. So one layer can be dedicated solely to consensus and another layer can be dedicated to execution.
So why is modular supposedly better than monolithic?
In the picture above, the visual on the left shows monolithic architecture, where both the consensus and execution functions are bundled into one layer and then a smart contract is built on top of that bundled layer. With monolithic architecture, you’re bound within one execution environment, limiting its potential to specialize and optimize for certain use cases.
The visual on the right shows Celestia’s modular architecture where you’ve separated the consensus and execution functions into separate layers, now the execution layer can exist as its own blockchain, where they can define their own state transition functions while also specializing and optimizing for certain use cases. This modular architecture or modular blockchain allows for better scalability, security and flexibility for any decentralized application built on it. We’ll get into the details of that soon.
In Ethereum, any individual with a laptop or computer can run a full Ethereum node, but it’s not really scalable because as more transactions are processed and the size of a block is increased, it becomes increasingly resource-intensive, meaning you would need better resources like higher hardware requirements i.e. a faster, more advanced computer, and not everyone has access or is able to acquire better resources, thus presenting a massive scalability issue.
This is what Celestia solves. Since it decouples (separates) consensus and execution, a full node in Celestia doesn’t have to execute every transaction that happens, it just reads every transaction as raw data. All it needs to make sure is that the transaction data in the block is available.
Now to verify data availability, the traditional approach would be to download all the data in the block. But then a problem occurs that as a block gets bigger, you’d face the same problem mentioned before where you'd then need better resources i.e. higher hardware requirements like a faster, more advanced computer and bigger data storage, thus limiting scalability.
To solve this scalability issue, Celestia came up with Data Availability Sampling. Rather than downloading the entire transaction data of a block, Data Availability Sampling allows you to sample small, random chunks of that block and once you’ve sampled enough of them, you’d have probabilistic certainty that the whole block is available.
Let’s use a simple example to help understand this concept better. Imagine you’re presented with 10 Gigabytes of data, instead of having to download the entire 10 Gigabytes to make sure that the data is there, with Data Availability Sampling you’d only need to download 100 Megabytes of that data and then you’ll know with 99.9% certainty that the entire data is available, thus solving the scalability issue of needing better resources as you’d only be downloading a small chunk or small percentage of that data instead of the entire thing.
What’s even crazier (in a good way) about how scalable this can be is that you can run Data Availability Sampling on a mobile phone, as shown by Mustafa, Celestia Labs’ CEO on a Twitter post:
In case you’re still having a hard time understanding Data Availability Sampling, I made this illustration that will hopefully help you get a better grasp of this concept.
Without Data Availability Sampling
Without Data Availability Sampling, in order to verify data availability, Pepo here would have to download the entire transaction data of a block. Initially it goes alright for him as his hardware (laptop) is able to keep up with the demands of processing the transaction data of the block. Now as Pepo wants to start scaling and starts processing more transactions, the size of the block increases. Pepo then realizes that his hardware is now unable to keep up with the increasing requirements necessary for scaling. Pepo is now unable to scale.
With Data Availability Sampling
With Data Availability Sampling, in order to verify data availability, Pepo only needs to sample a small, random chunk of data instead of the whole thing. The process is now not as hardware-intensive and requires fewer resources, thus Pepo is now able to scale.
Other than scalability, what reasons are there for dApps to launch on Celestia?
A few notable advantages for building on Celestia is the flexibility and sovereignty it offers. Let’s say a dApp is built on the Ethereum network, developers are then “stuck” within the limitations in its EVM (Ethereum Virtual Machine) which then restricts and limits the devs from being able to do certain things, limiting their original vision.
On the other hand, building on Celestia provides developers with greater flexibility where they can deploy not only on EVM, but also Cosmos SDK, Substrate and other execution environments like any IBC-enabled zone, freeing them from the limitations of a single chain.
I’m guessing this might be the most anticipated part of this write-up for some. The Celestia team has not specified when a token launch is happening, but don’t worry because a token is coming.
Despite no information as of yet of when a token for Celestia might launch, Nick White, COO of Celestia Labs, has given out some information on how and what the token might be like and what it might be used for in his interview with BigD_Senpai.
Celestia token can be used for staking and delegating, earning inflationary rewards and will also be what is used to pay for posting data on the blockchain and making it available.
For comparison, where in Ethereum you pay gas in ETH tokens to run computation on a shared EVM computer, in Celestia, using it’s native Celestia token, you pay per byte of data that you want to post. So the Celestia token for Celestia will be similar to what ETH is to Ethereum. They are also planning to implement a fee-burning mechanism similar to EIP-1559, bringing a deflationary aspect to the token. Nick also notes that as long as there is significant demand for secure, decentralized blockspace for people to run whatever decentralized computing they want, there will be significant demand for the Celestia token.
Modular blockchains have the potential to introduce a new paradigm shift in how decentralized applications are built as it improves upon the flaws and limitations of monolithic architecture, and in the face of this new paradigm shift, at the forefront of it, is Celestia.
Despite mainnet being months away from launching, Celestia has garnered a lot of attention and hype, and with its potential and what it promises to bring to the space, I’d say the attention and hype is warranted. Backed with a star-studded team and promising new innovations, if you’re a believer in the future of secure, decentralized blockspace and modular blockchains, I don’t see why you wouldn’t be a believer in Celestia.
This article was first published on 24 January 2022 in Substack. Minor edits were made for formatting, updated descriptions based on the Celestia website and any mentions of “Web3” were replaced. Thank you and goodbye.