Research Engineer - Accelerator Architecture

NVIDIA has been transforming computer graphics, PC gaming, and accelerated computing for more than 25 years. It’s a unique legacy of innovation that’s fueled by great technology—and amazing people. Today, we’re tapping into the unlimited potential of AI to define the next era of computing. An era in which our GPU acts as the brains of computers, robots, and self-driving cars that can understand the world. Doing what’s never been done before takes vision, innovation, and the world’s best talent. As an NVIDIAN, you’ll be immersed in a diverse, supportive environment where everyone is inspired to do their best work. Come join the team and see how you can make a lasting impact on the world. Here in the NVIDIA hardware accelerator team, we work on new architectures to support novel application areas and AI models. Our work includes researching the evolving landscape of AI applications and models, analyzing underlying model architectures, and building implementations for various NVIDIA hardware accelerators such as LPUs. We analyze mappings to existing and future accelerator systems, model performance, and work multi-functionally with the hardware design team on novel hardware features e.g. functional units, numeric modes, interconnect, system integration, etc. to unlock new application areas for NVIDIA accelerators. There are opportunities to participate in a wider range of R&D activities, either internally or externally with key NVIDIA partners. What you’ll be doing: AI application and model research Performance modeling Multi-functional work with hardware and software teams Next generation hardware architecture development Support internal and outward-facing R&D What we need to see: MSc or higher degree in CS/EE/CE/Mathematics or equivalent experience Minimum 2 years of relevant experience, preferably in AI models and applications Good foundation in computer science Knowledge of LLMs and other Gen AI applications Solid understanding of computer architecture and compute...