Description
We are seeking a Principal AI and ML Infra Software Engineer, GPU Clusters to join our Hardware Infrastructure team. As an Engineer, you will have a pivotal role in enhancing efficiency for our researchers by implementing progressions throughout the entire stack. Your main task will revolve around collaborating closely with customers to pinpoint and address infrastructure deficiencies, facilitating groundbreaking AI and ML research on GPU Clusters.
Key responsibilities include:
Engaging closely with our AI and ML research teams to discern their infrastructure requirements and barriers, converting those insights into actionable improvements.
Proactively identifying researcher efficiency bottlenecks and leading initiatives to systematically improve it. Driving the direction and long-term roadmaps for such initiatives.
Monitoring and optimising the performance of our infrastructure ensuring high availability, scalability, and efficient resource utilisation.
Helping define and improve important measures of AI researcher efficiency, ensuring that our actions are in line with measurable results.
Working closely with various teams, such as researchers, data engineers, and DevOps professionals, to develop a cohesive AI/ML infrastructure ecosystem.
Keeping up-to-date with the most recent developments in AI/ML technologies, frameworks, and successful strategies, and advocating for their integration within the organisation.
Requirements include:
BS or similar background in Computer Science or related area (or equivalent experience).
15+ years of demonstrated expertise in AI/ML and HPC tasks and systems.
Hands-on experience in using or operating High Performance Computing (HPC) grade infrastructure as well as in-depth knowledge of accelerated computing (e.g., GPU, custom silicon), storage (e.g., Lustre, GPFS, BeeGFS), scheduling & orchestration (e.g., Slurm, Kubernetes, LSF), high-speed networking (e.g., Infiniband, RoCE, Amazon EFA), and containers technologies (Docker, Enroot).
Capability in supervising and improving substantial distributed training operations using PyTorch (DDP, FSDP), NeMo, or JAX. Moreover, an in-depth understanding of AI/ML workflows, involving data processing, model training, and inference pipelines.
Proficiency in programming & scripting languages such as Python, Go, Bash, as well as familiarity with cloud computing platforms (e.g., AWS, GCP, Azure) in addition to experience with parallel computing frameworks and paradigms.
Dedication to ongoing learning and staying updated on new technologies and innovative methods in the AI/ML infrastructure sector.
Excellent communication and collaboration skills, with the ability to work effectively with teams and individuals of different backgrounds.