Responsibilities
Define the Performance Story Across the Stack Build benchmarking methodologies and performance frameworks that reflect real-world workloads from silicon to cloud. Partner with Engineering to validate results across chip architecture, compiler optimizations, runtime behavior, and cloud deployment. Analyze latency, throughput, scaling curves, memory behavior, and workload efficiency across LPU generations and model families. Maintain internal performance dashboards and source-of-truth materials that create alignment across Product, Marketing, Sales, and Executive Explain How the Stack Works and Why It Wins Translate engineering insights across hardware, software, and cloud operations into clear and credible explanations. Build technical narratives that articulate performance at each layer and show how the layers compound to create leadership. Requirements
Ten or more years in systems engineering, ML performance, computer architecture, or similar CS-driven roles. Deep understanding of inference workloads, profiling tools, optimization techniques, and benchmarking. Ability to explain full-stack performance, including silicon architecture, compiler behavior, runtime scheduling, and cloud deployment. Strong technical writing and communication skills with a high standard for clarity and accuracy. Experience building or maintaining benchmarking systems, analysis tools, or performance dashboards. Preferred
Benefits #J-18808-Ljbffr
Define the Performance Story Across the Stack Build benchmarking methodologies and performance frameworks that reflect real-world workloads from silicon to cloud. Partner with Engineering to validate results across chip architecture, compiler optimizations, runtime behavior, and cloud deployment. Analyze latency, throughput, scaling curves, memory behavior, and workload efficiency across LPU generations and model families. Maintain internal performance dashboards and source-of-truth materials that create alignment across Product, Marketing, Sales, and Executive Explain How the Stack Works and Why It Wins Translate engineering insights across hardware, software, and cloud operations into clear and credible explanations. Build technical narratives that articulate performance at each layer and show how the layers compound to create leadership. Requirements
Ten or more years in systems engineering, ML performance, computer architecture, or similar CS-driven roles. Deep understanding of inference workloads, profiling tools, optimization techniques, and benchmarking. Ability to explain full-stack performance, including silicon architecture, compiler behavior, runtime scheduling, and cloud deployment. Strong technical writing and communication skills with a high standard for clarity and accuracy. Experience building or maintaining benchmarking systems, analysis tools, or performance dashboards. Preferred
Benefits #J-18808-Ljbffr