Evaluate fabric performance as inference cluster size scales from 16 to 32 GPUs and beyond, identifying bottlenecks and building performance scaling models for state-of-the-art AI workloads.
Design and execute head-to-head performance comparisons against competing fabric switch solutions to produce data-driven differentiation evidence.
Test Infrastructure & Automation
Design, build, and maintain automated lab infrastructure including test execution pipelines, traffic generation tooling, and data collection and reporting systems.
Enable repeatable, high-quality, and scalable performance measurements across all hardware configurations, reducing manual effort and accelerating the test cycle.
Share infrastructure and playbooks with the Product Applications team to accelerate customer application development and issue resolution.
Cross-Functional Impact & Innovation
Partner closely with ASIC architecture, firmware, software, Product Definition, Product Applications, and Product Marketing teams to communicate findings, influence design decisions, and resolve performance-impacting issues.
Serve as a key technical resource in the early evaluation of new fabric architectures, interconnect technologies (UALink, PCIe Gen 6/Gen 7, Ethernet, UEC), and AI/ML communication paradigms.
Provide performance data, analysis, and live benchmark support for key customer engagements and industry events; produce clear, audience-appropriate performance reports, technical briefs, and marketing collateral, and maintain living documentation in Confluence.
Basic Qualifications
Bachelor's degree in Computer Engineering, Computer Science, Electrical Engineering, or a related technical field. We welcome both recent graduates with strong, directly relevant project, research, or internship experience and candidates with 2–5 years of industry experience in performance or systems engineering.
Hands-on experience running AI/ML workloads on GPU clusters — including benchmarking, performance analysis, and fine-tuning of workloads across clusters of GPUs or accelerators. This can come from industry, research, or substantial academic projects.
Demonstrated ability to debug and root-cause system-level performance issues across hardware, firmware, software, and network boundaries.
Excellent fundamental knowledge of compute algorithms, parallel algorithms, and AI/ML algorithms and workloads.
Strong working knowledge of computer systems, GPU systems, and datacenter networking — including PCIe and Ethernet fundamentals.
Working knowledge of GPU and CPU software stacks (e.g., CUDA, MPI, collective communication libraries, drivers, and OS-level performance tooling).
Proficiency in scripting and automation (e.g., Python) to build test pipelines and analyze large performance datasets.
Preferred Qualifications
MS or PhD in Computer Engineering, Computer Science, Electrical Engineering, or a related field.
Experience with scale-up fabrics and next-generation interconnects such as UALink, PCIe Gen 6/Gen 7, Ethernet, or UEC.
Deep understanding of modern inference and training workloads (LLMs, MoE, recommender systems) and their communication patterns.
Experience developing roofline models and competitive performance analyses for switching, networking, or accelerator silicon.
Sponsorship data shows employers' past filings in public US government records (DOL, USCIS, E-Verify) and is not a promise that any job will be sponsored. Immigration rules change. Verify your situation with your DSO or an immigration attorney.