Job Description
GPU Systems Engineer
A top-tier proprietary trading firm is seeking a GPU Systems Engineer to join their global infrastructure team. You will play a critical role in designing, optimizing, and maintaining the GPU compute environments that support advanced quantitative research and machine learning pipelines in a low-latency trading context.
Key Responsibilities:
- Design and maintain GPU-based systems for high-throughput, low-latency workloads across research and production environments.
- Collaborate with quants, researchers, and engineering teams to deliver high-performance GPU compute infrastructure for model training, simulation, and real-time inference.
- Optimize GPU system configuration at the hardware, OS, and driver level, including memory, I/O, and thermal performance.
- Automate deployment, monitoring, and maintenance of GPU clusters using modern IaC and DevOps tools.
- Evaluate new GPU hardware, interconnects (e.g., NVLink, Infiniband), and vendor driver stacks to inform infrastructure strategy.
Qualifications:
- Extensive experience managing and tuning GPU compute systems in production environments (NVIDIA preferred).
- Strong Linux systems engineering background, with deep understanding of kernel, drivers, and hardware interfaces.
- Experience with CUDA, NCCL, and GPU profiling/debugging tools (e.g., nvidia-smi, Nsight).
- Familiarity with orchestration tools and schedulers such as Kubernetes, Slurm, Airflow, or similar.
- Scripting and automation skills in Python, Bash, or equivalent.
- Knowledge of networking and storage optimization in high-throughput, distributed compute environments is a plus.
- Strong academic background in Computer Science, Electrical Engineering, or a related technical field.
Join a technology-first trading firm with a collaborative culture, flat hierarchy, and significant investment in research infrastructure. Enjoy top-of-market compensation and the opportunity to work on some of the most challenging and impactful systems in the industry.
Apply now for a confidential discussion.
Job Tags