7. Technical Architecture: The Sequencer & Batching
At the core of Frogment's performance lies the Sequencer, a high-performance off-chain component responsible for ordering transactions, executing state transitions, and compressing data for L1 settlement.
7.1 The Sequencer Lifecycle
The Sequencer is not just a server; it is a deterministic state machine designed for maximum throughput.
Ingestion (Mempool): Users submit signed transactions via RPC. Unlike L1 mempools which can be chaotic, the Frogment Sequencer orders these First-In-First-Out (FIFO) to prevent front-running.
Execution (Frogment VM): The transaction is processed against the current state. This happens in memory, allowing for TPS (Transactions Per Second) significantly higher than mainnet Solana.
Soft Finality: Once executed, the Sequencer returns a receipt to the user. From the user's perspective, the transaction is "done."
Batch Construction: The Sequencer groups thousands of processed transactions into a Batch.
7.2 Data Compression & Availability
To minimize the data footprint on Solana (and thus minimize costs), Frogment utilizes advanced compression techniques before posting to L1.
Transaction Stripping: Redundant data (like full public keys for recurring users) is replaced with short 4-byte Index IDs.
Zero-Byte Compression: Empty fields in transaction structs are omitted from the payload.
By maximizing (Batch Size) and minimizing Data Storage through compression, Frogment achieves fees that are a fraction of a cent ($< $0.0001).
7.3 Solana Data Settlement (The Anchor)
The compressed batch is wrapped in a standard Solana transaction and sent to the Frogment Anchor Program on Layer 1. This program:
Verifies the Sequencer's signature.
Stores the Data Root (Merkle Root of the batch) in a Solana Account (PDA).
Emits an event, signaling to Verifier Nodes that a new batch is available for inspection.
Last updated