Deep Learning Performance Architect

Job Title

Deep Learning Performance Architect

Role Summary

Join a deep learning performance team to model, measure, and optimize ML/DL workloads on GPU and accelerator-based systems. The role focuses on performance benchmarking, building and validating performance models, identifying bottlenecks, and proposing architecture and software optimizations for LLM/Generative AI workloads.

Experience Level

Mid-level. The posting indicates that 3+ years of relevant experience is a plus but not strictly required.

Responsibilities

Primary responsibilities include analyzing workload performance, creating projections, and driving actionable improvements across hardware and software.

• Benchmark and analyze performance of ML/DL workloads across GPU- and NPU-based architectures.
• Build, validate, and maintain performance models and deliver projections and insights for LLM/GenAI workloads on emerging architectures.
• Identify architecture, software, and system performance bottlenecks and propose actionable optimizations.
• Explore and evaluate new software and hardware capabilities and quantify application-level gains.
• Leverage AI agents to accelerate performance investigation and engineering workflows.

Requirements

Core qualifications and technical skills required or strongly preferred.

• Must-have: Strong background in computer architecture and system architecture design.
• Must-have: Familiarity with GPU or accelerator-based deep learning platforms and software stacks.
• Must-have: Knowledge of LLM/generative AI algorithms and kernel-level optimizations.
• Must-have: Experience with performance optimization and workload analysis for ML/DL systems.
• Must-have: Familiarity with machine learning and deep learning frameworks.
• Nice-to-have: Hands-on experience using AI agents to assist engineering workflows.

Education Requirements

BSc, MS, or PhD in Computer Science, Electrical Engineering, Mathematics, or a related technical discipline.

About the Company

Company: NVIDIA

Headquarters: Santa Clara, California, USA

NVIDIA is a global leader in accelerated computing, renowned for its innovative solutions in AI and digital twins that transform diverse industries. The company specializes in networking technologies, providing end-to-end InfiniBand and Ethernet solutions for servers and storage that optimize performance and scalability. NVIDIA serves sectors such as high-performance computing, enterprise data centers, and cloud computing, constantly reinventing its products and services to stay ahead in the market.