GTE Turbo is not a general-purpose blockchain. It is a trading-specific decentralized system designed from first principles to support perpetual futures, spot markets, tokenized equities, and real-world asset trading with verifiable state, ultra-low latency and institutional-grade reliability.Turbo separates settlement, margining, and execution into independent consensus clusters, enabling unprecedented throughput without sacrificing censorship-resistance or security.Turbo has three purpose-built layers:
Built on Zero, the Treasury Layer handles all deposits and withdrawals of assets in and out of the GTE protocol. It holds ERC20-compatible tokens and enforces withdrawal rules via authentication proofs against the Clearing House’s finalized QMDB state. Proof generation is lazily evaluated - users generate proofs when they want to withdraw, rather than requiring the protocol to produce them proactively. The Treasury may also implement protocol-wide health checks such as rate limiting and balance floors as defense-in-depth measures.
The Margin Engine is the authoritative source of truth for user margin state: account balances, subaccount margins, and positions. It processes all fills, validates margin requirements and account health, and executes permissionless liquidations when accounts breach solvency thresholds. Designed to be market-agnostic, its separation from trade production allows it to scale horizontally as the number of users grows.The Margin Engine is sharded by account - a trade between two users affects only their accounts, while other users’ state remains independent. This enables parallel processing across shards without coordination overhead. Validator design prioritizes decentralization and censorship resistance, with many validators distributed geographically running Minimmit: a consensus protocol operating that enables single-round finality through a nullify-by-contradiction mechanism.See more of our work on Minimmit.
Special Economic Zones (SEZs) handle order matching for specific trading pairs. Each is sharded by market, operating its own consensus and order book independently. A single static leader executes orders, with execution validated by a colocated group of matching engine validators for low latency finality. Our proprietary matching engine achieves sub-50ns write-ahead logging, enabling nanosecond order matching with price-time priority.SEZs are strategically colocated near price discovery centers for underlying spot assets - New Jersey for equities, Tokyo for crypto, Chicago for commodities. Market makers can colocate with SEZs for sub-millisecond execution.Learn more about SEZs.
We split up the architecture to hyper-optimize each piece, introducing a horizontally scalable blockchain and decentralized exchange.The result is an architecture that can handle 24/7 global trading without performance degradation.