Dynamic Shard Composition
- Why: Traditional static node participation in blockchains can lead to congestion and increased transaction costs.
- Example: Bitcoin’s fixed node structure results in slow transaction processing during peak times, leading to higher fees for users. PoR-BFT’s dynamic shard composition ensures efficient resource distribution by rotating nodes within each shard based on a pseudorandom seed generated by the CVSA.
Concurrent Transaction Application Within a Block
- Why: Throughput is bottlenecked by how many transactions a block can apply per second, not just by how many blocks the chain can produce.
- What ships today: Within a single block, the GCR pipeline partitions independent transactions into groups (via union-find over touched entity keys) and applies up to 8 groups concurrently while preserving sequential semantics within each group. See How is GCR Synced for the implementation.
Per-shard consensus itself runs sequentially — one consensus round per shard at a time — not multiple rounds in parallel within the same shard. The “parallel” gain in Phase 1 lives in transaction-application within a block (across independent entity groups), not in concurrent consensus rounds.