About the job
NVIDIA's Accelerated Computing team is a driving force behind the explosion of Machine Learning, Artificial Intelligence and High-Performance Computing. We are looking for a highly capable individual with a consistent track record in technology and the skills for GPU product definition for Data Center. We are a small, dynamic, and motivated team that defines the next generation of products for these high growth markets.
Responsibilities
Guide the architecture of the next-generation of GPUs through an intuitive and comprehensive grasp of how GPU architecture affects performance for datacenter applications, especially Large Language Models (LLMs)
Drive the discovery of opportunities for innovation in GPU, system, and data-center architecture by analyzing the latest data center workload trends, Deep Learning (DL) research, analyst reports, competitive landscape, and token economics
Find opportunities where we uniquely can address customer needs, and translate these into compelling GPU value proposition and product proposals
Distill sophisticated analyses into clear recommendations for both technical and non-technical audiences
Qualifications
Minimum
5+ years of total experience in technology with previous product management, AI related engineering, design or development experience highly valued
BS or MS or equivalent experience in engineering, computer science, or another technical field. MBA a plus.
Deep understanding of fundamentals of GPU architecture, Machine Learning, Deep Learning, and LLM architecture with ability to articulate relationship between application performance and GPU and data center architecture
Ability to develop intuitive models on the economics of data center workloads including data center total cost of operation and token revenues
Demonstrated ability to fully contribute to above areas within 3 months
Strong desire to learn, motivated to tackle complex problems and the ability to make sophisticated trade-offs
Preferred
2+ years direct experience in developing or deploying large scale GPU based AI applications, like LLMs, for training and inference
Ability to quickly develop intuitive, first-principles based models of Generative AI workload performance using GPU and system architecture (FLOPS, bandwidths, etc.)
Comfort and drive to constantly stay updated with the latest in deep learning research (academic papers) and industry news
Track record of managing multiple parallel efforts, collaborating with diverse teams, including performance engineers, hardware architects, and product managers