Member of Technical Staff, Research Engineer (GPU Performance)

Responsibilities

• Optimize training throughput across large GPU clusters using custom kernels and mixed-precision strategies
• Design and maintain distributed training infrastructure including tensor parallelism and FSDP
• Profile and accelerate inference pipelines for real-time multimodal generation

Requirements

• 4+ years of experience in systems engineering, ML infrastructure, or performance optimization
• Proficiency in GPU kernel development (CUDA, Triton, CUTLASS) and distributed systems
• Experience with ML framework internals such as PyTorch or JAX

About the Company

Runway is building AI to simulate the world through merging art and science, focusing on world models to accelerate progress in robotics, disease research, and scientific discovery.