We are looking for a first-class Deep Learning Performance architect to join in us to shape the performance analysis infrastructures for GPUs. We build cutting-edge analysis tools and visualization frameworks that empower engineers to optimize GPU performance for Deep Learning and HPC workloads—spanning pre-silicon architectural exploration to post-silicon validation and optimization. Your work will directly shape the tools that define how NVIDIA GPUs are analyzed, tuned, and scaled for next-gen AI systems, and impact the next-gen GPUs architectures.
What you’ll be doing:
- Architect Performance Tooling: Develop infrastructure tools/libraries for GPU performance analysis, visualization, and automated workflows used across GPU SW/HW development life cycle
- Unlock Architectural Insights: Analyze GPU workloads to identify bottlenecks and define new hardware profiling features that enhance perf debug and profiling capabilities.
- AI-Powered Automation: Build AI/ML-driven tools to automate performance analysis, generate perf optimization guidance, and improve user experience of profiling infrastructure.
- Cross-Stack Collaboration: Partner with kernel developers, system software teams, and hardware architects to co-design performance-centric solutions.
- End-to-End Optimization: Create benchmarks to validate performance improvements across AI/HPC workloads and present actionable insights.
What we need to see:
- BS/MS+ in relevant discipline (CS, EE, Math)
- Proficiency in C/C++ (performance-critical coding) and Python (automation/scripting, and AI/ML frameworks)
- Strong grasp of computer architecture (pipelines, memory hierarchies) and Operating System fundamentals
- Understand machine learning and data analysis basics, LLM techniques such as prompt engineering, fine-tuning, vector databases
- Experience with performance modeling, architecture simulation, profiling, and analysis.
- Self-starter who thrives in dynamic environments and manages competing priorities effectively.
Ways to stand out from the crowd:
- Experience with developing HW performance debugging and analysis tools
- Familiar with System Software Stack(like CUDA Driver), CUDA kernel optimization and understand GPU architecture
- Familiarity with GPU performance profiling tools like Nsight System, Nsight Compute
- Practical experience or projects demonstrating LLM-based code generation, automated data analysis, or workflow assistants. Prior experience with agentic LLM frameworks like Langchain and LLamaIndex.
- Full-Stack Versatility: Skills in JavaScript, SQL, or UI/UX design for tool interfaces.
Read Full Description