HPC Engineer

At IMC, technology is not a department; it's at the heart of everything we do. Developed in-house, our systems power world-class research and trading, enabling teams to make faster and better decisions every day. Our high-performance computing platforms sit at the core of this capability, supporting large-scale simulation, research, and machine learning workloads across the firm.
We're looking for an HPC Engineer to work closely with Researchers, Machine Learning Engineers, and Software Engineers to get the most out of our compute platforms. This role is about applying deep systems and parallel computing expertise to help users get the best possible performance from modern CPU and GPU-based infrastructure.
Your Core Responsibilities:
As an HPC Engineer, you'll act as a technical partner to our compute users, helping translate computational problems into efficient, scalable workloads. You'll work across teams to improve performance, throughput, and reliability, while shaping best practices for how HPC resources are used across IMC.

• Partner with Researchers, Quants, and MLEs to analyze, optimize, and scale HPC workloads across CPU and GPU platforms
• Support and optimize ML and simulation workloads, with a focus on performance, resource efficiency, and scalability
• Apply deep understanding of storage, networking, and scheduling systems to improve end-to-end workload performance
• Define and promote best practices for running large-scale workloads on shared HPC infrastructure
• Work with platform teams to influence system design decisions based on real workload behavior and performance data
• Implement and use performance monitoring, profiling, and benchmarking tools to drive continuous improvement
• Ensure workloads and platforms meet internal security and compliance standards

Your Skills and Experience

• Experience working with HPC or large-scale compute environments, with a strong focus on workload performance and optimization
• A solid grasp of parallel computing concepts and how to tune performance on CPUs and GPUs
• Hands-on experience with GPU acceleration (e.g. CUDA) and an understanding of how GPU workloads really behave
• Solid systems knowledge across Linux, storage and networking; enough to diagnose bottlenecks and guide users effectively
• Experience supporting or optimizing ML, simulation, or data-intensive workloads in shared compute environments
• Familiarity with containers and orchestration tools like Kubernetes is beneficial
• Programming experience is advantageous, particularly in Python and C++
• Strong communication skills and a genuine interest in helping others get better performance from complex systems
• A high degree of flexibility and adaptability: willing and able to deal with uncertainty and ambiguity in a rapidly evolving environment