What we already explored:

• Full replication
• KZG univariate with chunk sampling
• Multivariate polynomial commitment
• Greyhound polynomial commitment
• Blaze polynomial commitment
• Zoda with Hadamar and Tensor
• Frida
• Whir
• Semi-AVID

What we want to compare:

• Does it requires a trusted setup?
• What’s the size of the setup?
• Is it post-quantum?
• What’s the prover complexity for the disperser?
• What’s the size of a shard (without proof)? In percentage of the original data?
• What’s the size of the commitment(s) required to verify the shard?
• What’s the proof size of one shard?
• What’s the verifier complexity of one shard (DA node)?
• Is it supporting chunk sampling?