Dsquare is a highly specialized FX trading boutique. They are focused on consistently outperforming the market in short term trading. They use algorithmic trading models in the world’s biggest market-average daily volumes in foreign exchange to execute their strategies well in excess of $3 trillion.
In this excellent article, Jad Sarmo, Head of Technology @ DSquare company, shares his wonderful experience in building their high volume, low latency trading application. He explains the performance challenges their team faced and the solutions they came up with. He also showcases how their Java application was able to deliver better performance than C++. The same article also talks about how they used GCeasy tool to study and optimize garbage collection performance.