C++ Software Engineer (CUDA)

We are looking for a C++ software engineer to help us maximize the performance of GPU based processing within our systems. This includes tasks such as the efficient handling of real-time data streams, image processing and deep learning. Key Responsibilities

• Developing Algorithms in CUDA.
• Host - Device system management (memory management, data transfers, multi-GPU support).
• Deployment of ML models using TensorRT within a C++ application.
• Profiling GPU applications with Nsight Systems/Compute
• Configuration of GPU hardware to maximise performance.
• Make recommendations for suitable GPU hardware.
• Communicate the opportunities and challenges of GPU solutions to non-technical colleagues.

You would be working as part of our Software Engineering team and closely with our Machine Learning and Research teams. We have an informal and friendly company culture and a healthy work/life balance.

Required Skills, Knowledge and Expertise Candidates must have good general C++, software optimization skills, excellent CUDA, and be confident using CUDA libraries. Knowledge of networking, streaming, and video compression would also be beneficial but are not requirements.