Summary

Founding GPU Engineer at Fuse Energy will develop and optimise CUDA-based software for high-performance compute infrastructure, including multi-GPU clusters and energy-aware workload scheduling. The role combines low-level GPU performance engineering with real-time energy, grid, and data centre systems.

Responsibilities

• Design and optimise high-throughput cuda kernels
• Profile gpu compute and memory bottlenecks
• Correlate cluster power usage with grid signals
• Optimise multi-gpu and multi-node scaling
• Develop power capping and workload scheduling strategies
• Integrate custom kernels into ml pipelines
• Benchmark gpu performance against cpu baselines