For most of crypto's history, blockchains have been monolithic — one chain does everything. Execution, consensus, data availability, and settlement all happen on the same layer. Ethereum ETH$2,129ETH$2,12924h-0.08%7d-3.71%30d-15.52%1y-8.74%via Statility was built this way. Bitcoin BTC$72,798BTC$72,79824h-0.10%7d+0.41%30d-13.24%1y-21.76%via Statility was built this way. And as usage scaled, the limits of that design became painfully clear: congestion, high fees, and throughput ceilings that no amount of optimization could fully resolve.
The modular thesis offers a different architecture. Instead of one chain trying to handle every job, split those jobs across specialized layers. Let one chain handle execution, another guarantee data availability, another finalize settlement. Each layer optimizes for its specific function. The result, in theory, is a system that scales further, costs less, and avoids the compromises that monolithic chains inevitably make. Celestia TIA$0.3396TIA$0.339624h-0.03%7d-6.11%30d-7.96%1y-90.34%via Statility, EigenLayer EIGEN$0.1980EIGEN$0.198024h+0.00%7d-2.54%30d-23.50%1y-86.54%via Statility, and others are betting billions on this thesis being right.
What "Modular" Actually Means
Every blockchain performs four core functions. In a monolithic chain, all four run on a single network. In a modular stack, they are separated — or "unbundled" — into distinct layers that can be mixed and matched.
The Four Core Blockchain Functions
| Function | What It Does | Monolithic Example | Modular Example |
|---|---|---|---|
| Execution | Processes transactions and state changes | Ethereum L1 | Rollups (Arbitrum, Optimism, zkSync) |
| Data Availability | Ensures transaction data is accessible | Ethereum L1 (calldata/blobs) | Celestia, EigenDA, Avail |
| Consensus | Agrees on transaction ordering | Ethereum L1 validators | Shared sequencers, Celestia validators |
| Settlement | Finalizes state and resolves disputes | Ethereum L1 | Ethereum as settlement layer |
The idea is not new in computing. Operating systems unbundled from hardware decades ago. Cloud services separated compute, storage, and networking. The modular blockchain thesis applies the same principle: specialization beats generalization at scale.
Ethereum itself is already moving in this direction. Its rollup-centric roadmap effectively turns Ethereum into a settlement and data availability layer, with execution pushed to Layer 2 rollups. EIP-4844 (proto-danksharding) introduced blob transactions specifically to make Ethereum cheaper as a data availability layer. The modular movement is not anti-Ethereum — it is an extension of where Ethereum is already heading.
Celestia and the Data Availability Problem
Celestia TIA$0.3396TIA$0.339624h-0.03%7d-6.11%30d-7.96%1y-90.34%via Statility launched in late 2023 as the first blockchain designed exclusively for data availability. It does not execute smart contracts. It does not process user transactions. It does one thing: store and verify that transaction data has been published and is retrievable.
Why does this matter? Because data availability is the bottleneck that limits how cheap and fast rollups can be. When a rollup posts its data to Ethereum, it competes for block space with every other transaction on the network. Celestia offers a dedicated, purpose-built alternative. Rollups that use Celestia for data availability instead of Ethereum can reduce costs by orders of magnitude — from dollars to fractions of a cent per transaction.
Celestia uses a technique called data availability sampling (DAS). Instead of requiring every node to download all the data, light nodes can verify data availability by sampling random portions. This lets Celestia scale its data throughput as more nodes join the network, rather than being capped by what a single node can process.
The tradeoff: a rollup using Celestia for data availability instead of Ethereum inherits Celestia's security assumptions rather than Ethereum's. For many applications, that is perfectly acceptable. For others — particularly those handling billions in DeFi value — the security premium of Ethereum's validator set still matters.
EigenLayer: Restaking as Modular Security
EigenLayer EIGEN$0.1980EIGEN$0.198024h+0.00%7d-2.54%30d-23.50%1y-86.54%via Statility approaches modularity from a different angle. Instead of building a new chain, it lets new protocols borrow Ethereum's existing security. ETH stakers can "restake" their staked ETH to simultaneously secure additional services — called Actively Validated Services (AVSs) — on top of Ethereum.
This is modular security as a service . A new oracle network, data availability layer, or bridge does not need to bootstrap its own validator set from scratch. It can tap into Ethereum's $100+ billion in staked capital. EigenDA, the data availability service built on EigenLayer, competes directly with Celestia by offering Ethereum-aligned security.
The appeal is obvious: faster time to market for new protocols, stronger security from day one, and additional yield for ETH stakers. The risk is equally clear: restaking creates layered exposure. If a restaked validator is slashed on an AVS, the same ETH backing that validator's Ethereum staking could be affected. One failure cascading through multiple protocols is the scenario that keeps EigenLayer's critics up at night.
Avalanche Subnets and NEAR's Sharding
Not every modular approach requires separate chains. Avalanche AVAX$9.50AVAX$9.5024h+0.11%7d-3.36%30d-7.41%1y-54.00%via Statility and NEAR NEAR$1.30NEAR$1.3024h-0.59%7d+15.19%30d+12.50%1y-51.99%via Statility embed modular principles within their own ecosystems.
Avalanche subnets let developers launch custom blockchains with their own validator sets, consensus rules, and virtual machines — all within the Avalanche ecosystem. Each subnet is effectively a specialized execution layer. Gaming subnets can optimize for low latency. DeFi subnets can prioritize finality speed. Enterprise subnets can add permissioned access. The C-Chain remains the shared settlement environment, but execution is distributed.
NEAR takes a different path through sharding. Nightshade, NEAR's sharding design, splits the network into multiple shards that process transactions in parallel. Each shard handles a portion of the network's state and computation, but they share a single security model and can communicate with each other. It is modularity within a unified chain rather than across separate chains.
Both approaches try to capture the benefits of specialization without the fragmentation that comes from splitting across entirely separate networks.
Modular vs. Monolithic: The Solana Counter-Argument
Solana SOL$91.08SOL$91.0824h-0.14%7d-1.36%30d-16.86%1y-40.01%via Statility is the loudest voice on the other side of this debate. Its philosophy: build one chain that does everything, and make it fast enough that you do not need to split things apart.
Here is how the approaches compare on key dimensions:
Modular Stack vs. Monolithic Design
| Dimension | Modular (e.g., Celestia + Rollups) | Monolithic (e.g., Solana) |
|---|---|---|
| Scalability | Scales by adding specialized layers | Scales by optimizing single chain |
| Composability | Cross-layer bridging required | Native atomic composability |
| User Experience | Multiple chains, bridges, gas tokens | Single chain, single gas token |
| Security Model | Inherited from multiple layers | Unified validator set |
| Cost Structure | Very low (specialized DA layers) | Low but rises with congestion |
| Developer Complexity | High (choose and integrate layers) | Lower (one runtime, one environment) |
Solana's monolithic approach has real advantages. Every application on Solana can interact with every other application atomically, in the same block, without bridges or cross-chain messaging. A DEX trade can trigger a lending protocol action can trigger an NFT purchase — all in one transaction. In a modular world, that kind of composability requires cross-layer communication, which adds latency, cost, and failure modes.
The monolithic counterpoint also has data behind it. Let's see how these architecturally different assets have performed over the past year:
Indexed to 100 at start. Live data via Statility
Market performance does not validate architecture, but it reflects where capital and conviction are flowing. Solana's strong performance is partly driven by its vibrant application layer — something that becomes harder to build when execution is fragmented across many rollups and subnets.
The Risks of Going Modular
The modular thesis is intellectually elegant. But elegance in architecture does not automatically translate to superiority in practice.
- Complexity overhead. Developers building on a modular stack must choose an execution environment, a data availability layer, a settlement layer, and sometimes a shared sequencer. Each choice involves tradeoffs, integration work, and ongoing maintenance. Most developers would rather ship features than manage infrastructure.
- Fragmented liquidity. Every new rollup or subnet is an isolated liquidity pool. DeFi thrives on deep, concentrated liquidity. Splitting users and capital across dozens of execution environments works against this. Bridges help, but bridges are the most attacked infrastructure in crypto — over $2 billion has been lost to bridge exploits since 2021.
- Dependency chains. A modular application depends on its execution layer, its DA layer, its settlement layer, and any bridges or messaging protocols connecting them. If any link in that chain fails, degrades, or is compromised, the application suffers. Monolithic chains have a single point of dependency — which is simpler, even if that single point carries its own risks.
- User experience. Users should not need to understand what layer they are on. But in practice, modular architectures often leak complexity to the user: different gas tokens, bridging steps, finality times, and wallet configurations. The chains that win will be the ones where this complexity becomes invisible.
- Unproven at scale. Celestia launched barely a year ago. EigenLayer's restaking model is even newer. These systems have not been battle-tested through a full market cycle, a major exploit, or sustained congestion. The modular thesis has theoretical merit but limited empirical evidence.
Is Modularity the Future?
The honest answer: probably both approaches survive. The computing world did not converge on one architecture. Mainframes, microservices, monoliths, and serverless all coexist because different problems favor different designs.
High-frequency DeFi trading might thrive on Solana's monolithic composability. A gaming ecosystem with millions of low-value transactions might prefer a dedicated rollup with Celestia for cheap data availability. Enterprise applications might want Avalanche subnets with custom compliance rules. The question is not whether modular or monolithic wins — it is which use cases each serves best.
What is clear is that the modular stack is maturing rapidly. Celestia has proven that dedicated DA layers can slash costs for rollups. EigenLayer has demonstrated that shared security can bootstrap new protocols without starting from zero. Ethereum's own roadmap validates the core thesis. The infrastructure is being built.
The risk is over-engineering. Not every application needs a bespoke stack of specialized layers. Sometimes a fast monolithic chain that just works is the better answer. The modular thesis is powerful, but the winning version will be the one that hides its complexity from everyone except the developers who need it.
Modular blockchains are not replacing monolithic ones. They are expanding the design space. The best architecture is the one the user never has to think about.
Looking for crypto platforms, exchanges, and DeFi apps? Browse our curated directory:
