Description
At Snowflake, we are powering the era of the agentic enterprise.
We seek a Principal Performance Engineer to drive Snowflake's cloud infrastructure performance strategy. This is a high-autonomy, high-impact role for someone who can go deep on hardware internals while also translating that technical depth into pricing, capacity, and roadmap decisions across Snowflake.
Responsibilities
- Own new hardware evaluation and enablement. Lead evaluation of new AWS, Azure, and GCP instance types by building representative benchmark and evaluation solutions.
- Set performance strategy and architecture. Shape the team's technical roadmap, set direction on measurement methodology, and translate benchmark results into concrete rollout and adoption recommendations.
- Lead cross-functional and CSP partnerships. Work across engineering and Finance along with AWS, Azure, GCP to qualify new hardware and influence future cloud-instance designs.
- Characterize hardware performance at a deep level. Analyze CPU microarchitecture to identify workload-specific bottlenecks and drive resolution.
- Develop price/performance model to support new hardware transitions.
- Automate day-zero hardware readiness. Build automated performance-validation workflows and per-provider/per-instance scorecards so Snowflake can qualify and price new silicon at or near cloud-provider GA.
- Grow the performance engineering discipline. Mentor engineers, build out benchmarking capability, and drive company-wide performance education.
Requirements
- 12+ years of experience in performance engineering, systems engineering, or infrastructure engineering, with a track record of principal-level technical leadership.
- Deep expertise in cloud infrastructure performance across at least one major CSP, including instance types, pricing models, and capacity constraints.
- Strong grounding in hardware/systems fundamentals: CPU microarchitecture, memory bandwidth, I/O subsystems, and use of profiling tools to root-cause performance bottlenecks.
- Experience building or operating large-scale benchmarking systems and turning benchmark output into actionable optimization strategies and pricing/rollout decisions.
- Ability to build quantitative models connecting technical performance metrics to business outcomes.
- A track record of leading cross-functional, cross-company initiatives.
- Bias toward automation: builds scorecards, dashboards, and validation pipelines rather than relying on one-off analysis.
Nice to Have
- Experience with data warehouse / distributed database performance characteristics.
- Direct experience partnering with cloud providers or silicon vendors on early-access hardware programs.
- Experience building hardware-emulation systems.