Principal Performance Architect - ONTAP

NetApp•San Jose, CA

10d

About The Position

As a Performance Architect, you will lead the ONTAP performance team across hardware and software projects for ONTAP releases. You will work closely with architects across engineering development teams to set the performance direction, guide technical work, and drive the performance agenda, with the opportunity to influence hardware and software roadmaps. The role focuses on analyzing, benchmarking, and optimizing emerging workloads, including Generative and Agentic AI, and leading ONTAP’s AI benchmark strategy, including industry-standard submissions such as MLPerf Storage. It is part of Research and Development, which focuses on technology advances, competitive market and customer workload requirements, product cost and time-to-market, and product quality. The role owns ONTAP’s AI benchmark program and develops repeatable benchmark methods and workload models based on real customer AI pipelines. It involves working with GPU and AI ecosystem partners on reference architectures, certifications, and high-performance data paths, and translating results into technical reports, competitive briefs, and conference talks for product marketing and field teams.

Requirements

Typically 12–15 years of relevant experience.
Bachelor of Science, Master of Science, or PhD in Electrical Engineering or Computer Science, or equivalent experience.
In-depth knowledge of technology trends, architecture differentiation, competitive positioning, and workload optimizations.
Hands-on experience with AI/ML benchmarks and frameworks, and understanding of how AI training and inference stress storage through throughput, metadata, small random reads, and checkpoint write bursts.
Familiarity with GPU server architectures and high-performance data paths, including RDMA, NVMe-oF, GPUDirect Storage, and parallel file access.
Extensive knowledge and experience in computer operating systems, computer architecture and design, data structures, software design, and programming practices.
Systems programming experience in C and/or C++.
Strong scripting skills in Perl, Python (primarily within JupyterHub), and Shell.
Exceptional presentation and influencing skills.
Ability to make accurate work estimates and develop predictable plans.
Willingness to take on additional tasks and responsibilities that contribute positively to team, department, and company goals.

Nice To Haves

Knowledge of storage systems and file systems.
Experience publishing benchmark results or participating in industry groups such as MLCommons or SPEC.

Responsibilities

Evaluate and propose key architecture options to drive performance.
Assess ONTAP’s performance competitiveness and differentiation, including against competitors’ published AI benchmark results.
Optimize ONTAP performance for Generative and Agentic AI workloads across data ingest and preparation, training, checkpointing, inference, KV-cache, and RAG/vector search.
Own ONTAP’s AI benchmark program by planning, running, and publishing industry benchmark submissions such as MLPerf Storage and SPECstorage AI_IMAGE.
Define AI benchmark methods and workload models, including checkpoint/restore patterns and inference traces, that make real customer AI pipelines repeatable as tests.
Work with GPU and AI ecosystem partners on reference architectures, certifications, and high-performance data paths, including GPUDirect Storage, RDMA, NFS over RDMA, and pNFS.
Work closely with engineering and product management to drive the performance agenda.
Evaluate design alternatives for improved performance and prototype performance-enhancement opportunities.
Design, implement, execute, analyze, interpret, socialize, and apply storage-oriented performance measurements and results.
Create tools and automation for running AI benchmarks at GPU-cluster scale.
Turn benchmark results into technical reports, competitive briefs, and conference talks for product marketing and field teams.