Chronicle FIX – Socket Latency Optimisation

Benchmarks have a natural lifespan that can be improved with more modern hardware. This benchmark was performed on March 18, 2020. Please don’t hesitate to contact us if you have questions about how Chronicle FIX will perform today.

As part of an ongoing evaluation with one of our clients we have produced some benchmarks within Chronicle FIX to simulate a real world scenario. As with all our demos and benchmarks these are available within our GitHub repository.

The scenario we were given was – there are three processes; the client process, a proxy and an endpoint:

  • the client process simulates multiple clients and sends NewOrderSingles
  • the endpoint receives NewOrderSingles and responds with ExecutionReports
  • the proxy process sits in the middle and forwards NewOrderSingles from client to endpoint and ExecutionReports vice versa, and times their read to write times. (T2 – T1) for NewOrderSingles and (T4 – T3) for ExecutionReports

As the chart shows we managed to optimize the FIX Engine regardless of number of clients.

All timings in microseconds

 

Andrew Twigg

Subscribe to Our Newsletter

Featured Products

Trading

Chronicle
FIX

The fastest Java FIX Engine on the market with microseconds latency. Trusted by top-tier banks and funds with thousands of deployed sessions globally.

Read more >>