[[ZKP/ZKP Base Layer/ZKP Blockchain/Performance/Storage Layer|The Storage Layer]] operates within a decentralized network, comprising various node types that collectively ensure data availability and system resilience: - **Validators:** Nodes that [[ZKP/ZKP Base Layer/ZKP Blockchain/Technical Build Consensus Layer/Synchronization of PoI and PoSp in Consensus Rounds/Staking Power|stake coins]] (e.g., a hypothetical 1,000 ZKP coins as an example) participate in BABE block production and GRANDPA finality through Substrate's consensus mechanisms, earning rewards annually based on their combined PoI, PoSp, and staking contributions. - **Full Nodes:** Nodes storing 1TB of data through Substrate's storage backend, relaying transactions with a latency of 10ms via libp2p networking, and supporting the network's operation. These nodes maintain the complete blockchain state using Patricia Tries and coordinate off-chain storage through integrated workers, typically numbering around 1,000 in a moderate-sized deployment. - **Light Clients:** Resource-constrained nodes that synchronize with the network by downloading block headers and verifying proofs through Substrate's light client infrastructure, enabling participation without heavy storage demands while maintaining security guarantees. ### Peer-to-Peer Dynamics: - **Gossip Protocols:** Facilitate the dissemination of transactions in 50ms and blocks in 100ms across 1,000 nodes through Substrate's libp2p-based networking layer, ensuring efficient communication. - **Distributed Hash Tables:** Enable data lookups in 100ms through IPFS's Kademlia DHT integration with Substrate's off-chain workers, leveraging logarithmic scaling for scalability. - **Resilience:** Accommodates 10% daily node turnover and recovers from network partitions in 1 second through Substrate's adaptive networking and GRANDPA's partition tolerance mechanisms. These strategies ensure the [[ZKP/ZKP Base Layer/ZKP Blockchain/Performance/Storage Layer|Storage Layer]] can support the [[Overview of Use-Cases|growing demands of decentralized AI,]] from small-scale inference to [[ZKP/Data Marketplace/Federated Learning/Federated Learning within the Data Marketplace|large-scale training,]] with secure, private, and efficient data handling through Substrate's robust infrastructure. See also: [[ZKP/ZKP Base Layer/ZKP Blockchain/Exploring zk-Rollup for Future EVM Compatibility A Technical Analysis/zk-Rollups for Future EVM Compatibility|zk-Rollups for Future EVM Compatibility]]