Developer Technology Engineer - Energy

Company Overview

NVIDIA has been a pioneer in visual computing for over twenty years, revolutionizing computer graphics through the invention of the GPU. GPUs have since evolved to power deep learning and AI, enabling computers, robots, and autonomous vehicles to perceive and interpret their environments. NVIDIA seeks to expand its teams with top-tier engineers and computer scientists during this exciting period.

Role Overview

We are searching for an expert Compute Developer Technology Engineer (DevTech) specializing in the Energy sector to accelerate simulation and AI workflows on NVIDIA platforms. The role emphasizes CUDA performance optimization for complex applications like seismic imaging, reservoir simulation, power grid modeling, and related HPC/AI workflows. Collaboration with customers, partners, and internal NVIDIA teams will be key to delivering scalable multi-GPU and multi-node performance enhancements.

Key Responsibilities

• Profiling and analyzing GPU-accelerated applications, focusing on CUDA kernels, memory access, concurrency, and overall throughput.
• Enhancing performance by optimizing CUDA C++ kernels, tuning launch configurations, exploiting memory hierarchies, and utilizing streams/events effectively.
• Leveraging GPU libraries such as cuBLAS, cuFFT, cuSPARSE, cuSOLVER, and NCCL to maximize efficiency.
• Enabling multi-GPU and multi-node scalability using MPI alongside NCCL, optimally overlapping CPU and GPU workloads and refining communication patterns.
• Creating reproducible benchmarks, producing thorough performance analyses, and delivering tuning advice, including scaling curve evaluations.
• Building and maintaining reference implementations, examples, and customer code patches to ensure high performance and code portability.
• Supporting customer engagements from proof-of-concept to production stages by debugging, ensuring correctness, advising deployment best practices such as containerization, scheduler usage, and cluster management.
• Collaborating internally to document issues, verify fixes, and influence NVIDIA’s product roadmap based on energy domain customer feedback.
• Developing internal reusable libraries and code bases that contribute to future NVIDIA product offerings.

Candidate Requirements

• A BS/MS degree or equivalent experience in Computer Science, Computer Engineering, Electrical Engineering, Physics, Applied Mathematics, or related disciplines.
• Proficiency in C/C++ and Python programming within a Linux environment.
• Demonstrated hands-on experience in CUDA programming coupled with GPU performance optimization expertise.
• Exposure to profiling and debugging tools such as NVIDIA Nsight Systems and Nsight Compute or similar.
• Strong understanding of parallel computing principles, including vectorization, threading, NUMA, and memory bandwidth/latency considerations.
• Excellent communication skills, capable of conveying technical insights clearly to both technical and non-technical audiences.
• At least five years of relevant experience in GPU and HPC performance tuning, showing a proven history of delivering measurable speedups and scaling improvements.

Preferred Qualifications

• Experience leading performance reviews and creating reusable playbooks or reference designs.
• HPC competencies with MPI, distributed systems, and performance tuning across multiple nodes.
• Domain expertise in Energy HPC including seismic data processing pipelines (RTM, FWI), FFT/stencil/linear algebra intensive computations, reservoir simulations involving sparse solvers, preconditioning, and domain decomposition techniques.
• Knowledge of power grid simulation tasks, transient stability analysis, and optimization workflows.
• Familiarity with continuous integration and performance regression testing as well as containerized workflows using Docker or Apptainer and job schedulers such as Slurm.
• Understanding of AI workflows integrating with simulations, including data preparation, training, inference, and pipeline optimization.

Additional Information

NVIDIA is one of the most respected technology employers globally, attracting innovative and independent professionals. If you exhibit creativity and autonomy, this opportunity is ideal for you.

Job Reference: JR2018521