This role pays more than
99%
of similar roles. Most pay
$171,500–$246,150
— the shaded band above.
At the midpoint, this role pays about
$352k
versus about
$209k
for comparable roles.
Based on 239 similar postings.
Employer
About Nvidia
Nvidia is a leading designer of graphics processing units (GPUs) and system-on-chip units, powering gaming, professional visualization, data centers, and artificial intelligence workloads. Industry: Semiconductors & AI Computing
Nvidia currently has
967 open roles
on FindRole.
Listed pay typically runs
$168,000–$270,250
across 950 roles with salary data.
NVIDIA is hiring a Principal Engineer to lead the optimization of large-scale AI training and post-training workloads on its advanced hardware and software platforms. This role involves analyzing and enhancing frontier-scale LLM workloads running on thousands of GPUs, driving improvements in frameworks like PyTorch, JAX, NeMo, and NeMo RL, and shaping future NVIDIA GPU, system, and software roadmaps based on real-world insights. The ideal candidate will have extensive experience in large-scale AI training systems, GPU performance optimization, distributed systems, and high-performance computing, with a deep understanding of GPU architecture from individual accelerators to datacenter-scale systems. They should be proficient in using profiling, tracing, benchmarking tools, and possess strong technical leadership skills to influence multi-functional decisions across NVIDIA’s teams. This role offers the chance to collaborate on cutting-edge AI projects that impact the future of computing and social progress.
Lead end-to-end performance analysis and optimization of large-scale LLM training on NVIDIA platforms.
Identify and eliminate bottlenecks in compute, memory, communication, and scheduling for AI workloads.
Develop software tools and benchmarks to enhance efficiency and developer productivity across AI stacks.
Guide future GPU and system architecture decisions with insights from workload characterizations and simulations.
Serve as a technical expert for AI training performance, collaborating with cross-functional teams at NVIDIA.
What we're looking for
MS or PhD in Computer Science, Electrical Engineering, or related field with 12+ years of relevant experience.
Proven technical impact in large-scale AI training systems, GPU optimization, distributed systems, HPC, ML frameworks, compilers/runtimes, or hardware/software co-design.
Deep hands-on expertise in analyzing and optimizing performance of large-scale deep learning workloads, especially transformer-based models.
Strong understanding of GPU and AI accelerator architecture from individual accelerators to datacenter-scale systems.
Experience with distributed training techniques including various parallelism strategies and mixed precision training.
Extensive use of profiling, tracing, benchmarking, and modeling tools to diagnose complex bottlenecks and drive performance improvements.
Excellent communication and technical leadership skills to influence multi-functional decisions across teams.
Python
C++
CUDA
PyTorch
JAX
GPU
MLPerf
NVIDIA
Deep Learning
Computer Architecture
Performance Modelling
Automation Tools
System Simulators
Cloud Services
Data Centers