Principal High-Performance LLM Training Engineer
$272k - $431k • Remote • US, CA, Santa Clara
Posted 1d ago
Job Location
US, CA, Santa Clara
Tech Stack
Remote Work Policy
Fully remote
Categories
LLM Engineer
About the job
NVIDIA is seeking a Principal Engineer to drive the performance of large-scale AI training and post-training workloads across NVIDIA’s full hardware and software stack. This role sits at the intersection of distributed training, GPU architecture, systems software, deep learning frameworks, and performance engineering. You will analyze and optimize frontier-scale LLM workloads running on thousands of GPUs, drive improvements across frameworks such as PyTorch, JAX, NeMo, and NeMo RL, and use insights from real workloads to help shape future NVIDIA GPU, system, and software roadmaps. We are looking for a deeply technical leader who can operate across abstraction layers: from application-level training behavior to framework/runtime internals, CUDA libraries, communication collectives, memory systems, networking, and GPU architecture. At this level, success means both directly improving performance directly as well as setting technical direction, raising the bar for the organization, and influencing multi-functional decisions across NVIDIA.
Responsibilities
- Lead end-to-end performance analysis and optimization of LLM pre-training and post-training workloads on NVIDIA hardware and software.
- Identify and remove bottlenecks across compute, memory, communication, scheduling, parallelism, kernels, frameworks, and system scaling to improve workload performance.
- Develop production-quality software, tools, models, benchmarks, and analysis infrastructure to enhance training performance, efficiency, and developer velocity.
- Build and refine performance models and workload characterizations to guide future GPU, networking, system, and software architecture decisions.
- Serve as a technical authority for AI training performance, collaborating with GPU architecture, systems, CUDA libraries, compilers, networking, frameworks, and product management teams.
- Translate workload insights into hardware and software recommendations to improve performance and efficiency across the AI ecosystem.
- Mentor and provide technical leadership to engineers, establishing best practices for large-scale AI performance analysis and optimization.
Requirements
- MS, PhD, or equivalent experience in Computer Science, Electrical Engineering, Computer Engineering, or related field, with 12+ years of relevant work or research experience.
- Demonstrated principal-level technical impact in large-scale AI training systems, GPU performance optimization, distributed systems, high-performance computing, ML frameworks, compilers/runtimes, or hardware/software co-design.
- Deep hands-on experience analyzing and optimizing performance of large-scale deep learning workloads, especially transformer-based models, LLM pre-training, reinforcement learning, fine-tuning, or other post-training workloads.
- Strong understanding of GPU and AI accelerator architecture from individual accelerators to datacenter-scale systems.
- Experience with distributed training techniques including data parallelism, tensor parallelism, pipeline parallelism, expert parallelism, sequence parallelism, activation checkpointing, mixed precision training, and communication/computation overlap.
- Proven track record using profiling, tracing, benchmarking, and performance modeling tools to diagnose complex bottlenecks and drive improvements.
- Excellent communication and technical leadership skills, with the ability to influence architecture and software decisions across multiple teams.
Benefits
- Equity
- Benefits