Sharding and Parallel Execution

Sharding is how you scale horizontally without blowing up node requirements. Split the state, split the processing, and keep shared security. It sounds clean. The single-shard takeover problem is why it isn’t. This chapter works through that problem and its solutions—committee rotation, erasure coding, and data availability sampling—then covers execution vs. DA sharding, cross-shard communication trade-offs, how Ethereum’s sharding story went from Eth 2.0 to EIP-4844 to full danksharding, and what Zilliqa, NEAR, and Polkadot each chose differently. Here’s what it covers: