Architecture

Co-location on a chip

At a typical exchange, market data is published from the exchange and consumed on separate trading servers. HFTs respond by sending orders back across the network. Tick-to-trade latency includes the publication network latency, the decision latency, and the order send network latency.

The PAX λ API hosts participant decision logic on the same chip as the matching engine. Market data is available to λ programs within nanoseconds of publication, without network transit between separate machines. A transaction at the most evolved traditional exchanges takes approximately 30 microseconds; at high-volume crypto exchanges, several milliseconds. A transaction at PAX takes approximately 30 nanoseconds.

A λ program is a predicate paired with an underlying limit order. When market conditions match the predicate, the order fires from the λ book into the matching engine, on the same silicon.