Nomos DA v1

• Fixed size chunks
• Number of Columns is fixed (related to the number of subnetworks)
• Number of Rows is variable. As data grows, the number of rows can grows
• Good for small datasets but inefficient for large datasets (e.g., encoding and proof generation become slow beyond ~1 MB)

Nomos DA v2

• Need square matrix ($n \times n$)
• Fixed matrix dimensions (to continue the existing subnetwork structure)
• Chunk size adjusts based on the total blob size and fixed matrix dimensions.
• No need to generate proofs (encoded rows and columns are also proofs)
• Very efficient encoding time