C++ Developer (GPU)

C++ Developer (GPU) works with the C++ team and researchers to identify opportunities to speed up the calculation (either in terms of latency or throughput) using GPU. You will work with a wide range of applications from different teams: it could be a parallel data processing pipeline from another software engineer team, or numerical computation for the researcher team. You will also work with the infrastructure team to build up a more systematic approach to allow other teams to utilize GPU. Responsibilities: Optimizing the performance of various components of the trading system Maintaining and enhancing backtest simulator Exploring big-data infrastructure for research Exploring execution algorithm research

Familiar with modern C++ (11, 14, 17 or 20) GPU programming experience (NVIDIA CUDA framework: basic understanding of blocks/threads launch config, shared memory, latency hiding etc. Strong background in algorithm, data structures and/or operating systems-related topics (memory management, scheduler, I/O etc.) Ability to quickly learn new concepts or technologies Ability to write sizable applications and manage complexity Linux experience Effective communicator and decision maker Knowledge of automated trading system Basic knowledge in operating system concepts: threads, processes, context switch, virtual memory etc. Basic knowledge in applied computer architecture concepts: SIMD, superscalar execution, memory/instruction latency vs bandwidth.