Latency
In trading systems, latency is the end-to-end delay between an external event (market data update, order arrival) and the system's response (order submission, quote update). Latency is additive across network transmission, serialization/deserialization, application logic, and queue delays. In crypto markets, latency matters differently depending on the venue: centralized exchanges using CLOBs reward low-latency participants in the classic HFT style (co-location, FPGA, kernel bypass networking), while on-chain AMMs and PBS auctions introduce a different latency regime where the fixed block time (e.g., 12 seconds on Ethereum) dominates. In the Ethereum MEV context, latency to the builder matters — faster propagation of order flow to builders allows faster inclusion and better MEV extraction. Cross-chain latency arbitrage exploits price differences arising from variable block times and bridge finality delays across networks. Trading infrastructure optimization in crypto spans the full spectrum from hardware-level (smart NICs, deterministic latency measurement) to protocol-level (preconfirmations, based sequencing).