Summary

In this chapter, we covered a lot of modern C++ 14, 17, and 20 features for multi-threaded and ultra-low latency applications that deal with shared memory interactions. We also covered static analysis for application development. Finally, we discussed runtime performance optimization techniques that move as much decision, code evaluation, and code branching away from runtime into compile time as possible. The contents of this chapter address the major design and development decisions and techniques critical to the development and performance of C++ HFT applications.

We also saw how to build an HFT system for a small hedge fund specialized in FX. In the next chapter, we will learn about the usage of Java in an ultra-low-latency system such as an HFT system.