Avail

Avail is a modular blockchain network that provides a specialized data availability (DA) layer, ensuring that transaction data for other blockchains is published and verifiably accessible without requiring full nodes to download entire blocks. This core function is critical for the security of optimistic rollups and validiums, which rely on the availability of their data to allow fraud proofs or to reconstruct their state. By decoupling data availability from execution, Avail enables higher scalability and lower costs for layer-2 solutions and sovereign chains built on top of it.

The network operates using a Nominated Proof-of-Stake (NPoS) consensus mechanism and leverages advanced cryptographic techniques like KZG polynomial commitments and data availability sampling (DAS). With DAS, light clients can randomly sample small portions of a block to verify with high probability that all data is available, a process that is both secure and highly efficient. This architecture allows Avail to support a high throughput of data, measured in megabytes per second, making it a robust foundation for a modular blockchain stack.

Avail's primary product is its Avail DA layer, but its roadmap includes Avail Nexus, a unification layer for cross-rollup interoperability, and Avail Fusion, a mechanism for securing the network with multiple crypto-assets like ETH and BTC. By providing these modular components, Avail aims to solve the fragmentation and scalability challenges within the broader ecosystem, positioning itself as essential infrastructure for developers building scalable and sovereign application-specific chains.

Avail operates as a specialized data availability (DA) layer, separating the critical functions of consensus and data publication from transaction execution. Its core mechanism uses data availability sampling (DAS), where light clients can verify that all transaction data for a block is published and accessible by randomly sampling small, random portions of the block. This allows nodes to confirm data availability with high confidence without downloading the entire block, enabling secure and scalable blockchain scaling. The network achieves consensus on the ordering and availability of data, not on the validity of state transitions, which is delegated to execution layers like rollups.

The architecture is built around three primary components: the Avail DA layer, Avail Nexus, and Avail Fusion. The DA layer is the foundational chain where rollups and other chains post their transaction data, secured by a Nominated Proof-of-Stake (NPoS) consensus mechanism using the BABE and GRANDPA protocols from the Polkadot SDK. Validity proofs, such as KZG polynomial commitments, are used to create compact cryptographic proofs that the data is available and correctly encoded, allowing for efficient verification by light clients and other chains.

Avail Nexus acts as a cross-chain coordination hub, functioning as a verification layer that sits atop the DA layer. It resolves cross-domain proofs and messages, providing a unified bridge for rollups and sovereign chains to interoperate securely. Avail Fusion introduces a novel cryptoeconomic security model where multiple assets (like ETH and BTC) can be staked to back the network's security, moving beyond a single native token to create a more robust and decentralized economic base. This modular stack allows execution layers to leverage Avail for scalable, secure data publishing while maintaining sovereignty over their own logic and governance.

Data Availability Sampling (DAS) is a protocol that allows light clients to probabilistically verify that all data for a block is published and accessible without downloading the entire dataset. By performing multiple random queries for small pieces of data, a client can achieve high statistical confidence that the data is available. This is critical for preventing data withholding attacks, where a malicious block producer creates a valid block but withholds some data, making it impossible for others to reconstruct the block's state and detect fraud.

The efficiency of DAS relies on representing block data with an erasure-coded data structure, typically using a 2D Reed-Solomon encoding scheme. This process expands the original data with redundant parity data, creating data "shards." The key property is that the entire dataset can be reconstructed from any sufficient subset of these shards (e.g., 50% in a 2x expansion). This redundancy allows sampling to be effective; even if some shards are missing, the data remains recoverable from the remaining ones, and sampling can detect their absence.

To commit to this erasure-coded data in a way that allows for efficient sampling proofs, systems like Avail employ KZG polynomial commitments (also known as Kate commitments). A KZG commitment is a constant-sized cryptographic fingerprint that binds a prover to a polynomial. The block data is interpolated into a polynomial, and the commitment is published. Crucially, the prover can generate a tiny, constant-sized proof for any individual data shard (a polynomial evaluation) that can be verified against the single master commitment, proving the shard's correctness without revealing the entire polynomial.

The combination of these two techniques creates a powerful security model. Light clients use DAS to randomly select shard indices, then request KZG evaluation proofs for those specific shards from the network. By verifying these proofs against the block's KZG commitment, the client is assured that each sampled shard is correct and part of the committed data. After a sufficient number of successful random samples, the client accepts that the entire data blob is available with overwhelming probability.

This architecture enables the separation of consensus and execution. A layer like Avail can provide a robust, scalable data availability layer where execution layers (rollups) simply post their transaction data. The execution layer validators only need to trust that the data is available via these cryptographic and statistical guarantees, allowing them to focus solely on processing transactions. This modular design is a cornerstone of modern blockchain scalability, reducing the resource burden on any single participant.