Senior Distributed AI & HPC Systems Architect

NVIDIA AI; Zürich; Vor Ort; CHF 190.000 - 260.000; Vollzeit

Benefits dieser Stelle

Competitive compensation

Comprehensive benefits package

Zusammenfassung

NVIDIA Zurich is seeking a PhD-level systems engineer to help build scalable software for distributed AI training and inference. You will collaborate with AI framework teams (TensorFlow, PyTorch, JAX) and internal platforms to optimize end-to-end performance across GPUs, DPUs, and interconnects.

The role requires deep expertise in C++, Python, CUDA, and experience delivering production-grade, high-throughput systems.

Qualifikationen

• Ph.D., or equivalent industry experience in CS or related field.
• 5+ years of experience in systems programming, parallel or distributed computing.
• Strong programming background in C++, Python, and CUDA with production-grade performance.
• Extensive hands-on experience with AI frameworks (PyTorch, TensorFlow, JAX).
• Demonstrated ability to reason across software stacks, hardware capabilities, and bottlenecks.

Aufgaben

• Build and evolve scalable architectures for distributed AI training and inference, optimizing throughput, latency, resiliency, and memory efficiency.
• Develop and evaluate next-generation communication and runtime capabilities in libraries like NCCL, UCX, and UCC.
• Collaborate with AI framework teams and internal platforms to build integrations and improve end-to-end performance and reliability.
• Work on hardware and system-level features across GPUs, DPUs, and interconnects to speed data movement.
• Drive runtime innovations and practical implementations across software and hardware stack.

Kenntnisse

C++

Python

CUDA

Ausbildung

Ph.D. or equivalent

Tools

NCCL

UCX

UCC

Jobbeschreibung

NVIDIA Zurich is seeking a PhD-level systems engineer to help build scalable software for distributed AI training and inference. You will collaborate with AI framework teams (TensorFlow, PyTorch, JAX) and internal platforms to optimize end-to-end performance across GPUs, DPUs, and interconnects.

The role requires deep expertise in C++, Python, CUDA, and experience delivering production-grade, high-throughput systems.