Used Tools & Technologies
Not specified
Required Skills & Competences ?
Marketing @ 3 Grafana @ 3 Kubernetes @ 3 Linux @ 3 Prometheus @ 3 DevOps @ 3 Python @ 3 CI/CD @ 3 TensorFlow @ 3 Leadership @ 3 Bash @ 3 Communication @ 5 Networking @ 3 Product Management @ 3 PyTorch @ 3 CUDA @ 3 GPU @ 3Details
Intelligent machines powered by Artificial Intelligence are rapidly moving from science fiction to reality. NVIDIA's GPU deep learning and visual computing technologies extend into datacenters, mobile devices, and cars. This role on the Ethernet Networking team focuses on demonstrating and communicating the performance leadership of NVIDIA's Spectrum‑X Ethernet platform for modern AI infrastructure.
Responsibilities
- Design and execute performance benchmarks using industry-standard tools (e.g., MLPerf, UCX, NCCL, CloudAI) and customer-representative AI workloads on GPU clusters.
- Drive performance characterization of complex training and inference workloads on AI supercomputers; develop metrics to isolate bottlenecks and guide optimization across the silicon-to-software stack.
- Translate benchmark data and technical insights into high-impact marketing assets and sales enablement materials (white papers, blogs, presentations).
- Collaborate with Product Management, ASIC and Software Architecture, and Sales to provide feedback on product features and ensure performance results are accurate and impactful.
Requirements
- B.Sc. in Computer Science, Software Engineering, or equivalent experience.
- 5+ years of experience benchmarking and analyzing high-performance networking solutions, including RDMA, MPI, and large-scale collective communication frameworks.
- Hands-on expertise testing and benchmarking deep learning workloads on NVIDIA GPUs with CUDA, TensorFlow, and PyTorch, focused on distributed training and inference performance over NCCL, RoCE, and RDMA.
- Proficiency in Performance Analysis methodologies and techniques.
- Understanding of Ethernet and high-performance networking.
- Programming experience with Python, Bash, and C.
- Experience with distributed job orchestration (Slurm, Kubernetes).
- Experience with Linux distributions.
- Strong analytical, problem-solving, and fast self-learning capabilities.
- In-depth knowledge and experience with AI workloads and benchmarking for large-scale distributed training/inference systems.
Preferred / Ways to Stand Out
- Strong performance analysis skills and methodologies using modern tools.
- Deep knowledge of AI/data center Ethernet network protocols and best practices (Clos fabrics, BGP, VXLAN).
- Hands-on experience with automation, CI/CD pipelines, and DevOps practices.
- Expertise in AI fabrics telemetry, metrics capture and analysis, and telemetry tools such as Prometheus and Grafana.
- Broad system knowledge (Intel/AMD/ARM CPUs, NVIDIA GPUs, NICs, memory, PCI).
Compensation & Benefits
- Base salary ranges by level (location and experience dependent):
- Level 3: 128,000 USD - 201,250 USD
- Level 4: 148,000 USD - 235,750 USD
- Eligible for equity and benefits. (Link to NVIDIA benefits provided in the original posting.)
Additional Information
- Applications accepted at least until January 15, 2026.
- NVIDIA uses AI tools in its recruiting processes.
- NVIDIA is an equal opportunity employer committed to diversity and non-discrimination.