High-Performance C++

---

Financial institutions and especially front-office banking are dependent on high-speed mathematical calculations to make informed trading decisions, manage risks, engage in high-frequency and algorithmic trading, and process large amounts of data efficiently. The ability to perform these calculations quickly and accurately is a competitive advantage and can lead to cost savings, better financial outcomes and open-up new business opportunities. Understanding modern computer hardware can allow a 10x performance improvement.

• Innovation: High-speed mathematical finance calculations enable banks to develop innovative products and services that require complex calculations and analysis. This can help them stay ahead of the competition and meet the evolving needs of their customers.
• Scalability: Faster calculations enable financial institutions to scale their operations and grow their business and increase their revenue.
• Risk management: Financial institutions need to manage risks associated with their investments and trading activities. High-speed calculations are necessary to perform real-time risk assessments, calculate overnight sensitivities and make informed risk management decisions.
• High frequency/e-trading: High-frequency trading (HFT) and e-trading relies on extremely fast calculations. HFT firms use complex algorithms that require fast mathematical calculations to identify profitable trading opportunities.
• Algorithmic trading: Algorithmic trading involves using computer programs to execute trades based on pre-defined rules. These rules often involve complex mathematical calculations that need to be performed quickly and accurately.
• Opportunity: If calculations can be performed 10x or 100x faster, what new business opportunities will be possible

The course covers

• Stable Performance: Considerations for how to achieve performance improvements without harming code quality, readability and maintainability.
• Memory: The impact of memory use and organisation on performance including virtual memory, cache hierarchy, cache-friendly software, impact of heap and stack memory.
• Hardware: How to code to benefit from modern hardware design including super-scalar (vectorisation) and SIMD coding, including openMP. Choices on multithreading and grid-based computing.
• Helping Compiler: How to design code to convey intent to the compiler, how to use C++ language and extensions to assist the compiler.
• Algorithm Design: How to design or modify algorithms to work well with modern hardware.  
• New Code design: Design considerations for green-field code development
• Legacy Code/Optimisation: How to tackle performance tuning legacy code. Selection of what to tune, profiling tools.

Experienced quant and quant-dev teams with a C++. A numerate background is a benefit for some examples.

Course contents and duration can be modified to align with additional client needs. 

Delivery can be on-site, remote or in recorded form.