The transition from Nomos DA v1 to Nomos DA v2 introduces ZODA to enhance efficiency, reduce proving overhead, and improve scalability. Unlike v1, which relies on RS encoding + KZG commitments, v2 plans to integrate a structured tensor-based encoding with RS expansion and Merkle commitments for integrity verification. This shift necessitates modifications to both encoders and verifiers, ensuring seamless integration of ZODA while maintaining compatibility with Nomos DA architecture.

The followings outline the required modifications to integrate ZODA into Nomos DA:

Data Representation and Encoding

• The data is structured as an $n \times$ n matrix.
• Each row and column is RS with a rate of $1/2$.
• This follows the Nomos v1 encoding model, but now both rows and columns will be sampled, which aligns with ZODA’s requirements.
• Chunk size is fixed in v1, but now in v2 it is variable.

Data Distribution

• Subnetwork assignment remains the same as Nomos v1, meaning that different nodes will hold different portions of the encoded matrix.
• However, instead of only relying on random sampling of columns, ZODA ensures full row and column inclusion in the network for proper verification.
• Enough rows and columns must be sent to subnetworks.

Encoder Limits

NomosDA v1 supports blob sizes from 32 KB to 1024 KB, primarily constrained by the computational resources of executors and bandwidth limitations. However, with the integration of ZODA in Nomos v2, these limits can be expanded.

Sub-Protocols

General idea of the sub-protocols (dispersal, replication, sampling, reconstruction) mentioned in the NomosDA Network Specification will remain the same in Nomos V2. However, due to changes in the data structure and cryptographic structure, the implementation will require updates.