Sr. Performance Modeling Architect

Job Title

Senior Performance Modeling Architect

Role Summary

Design and validate performance models to influence microarchitecture and system-level decisions for high-performance RISC-V CPUs targeting AI and HPC workloads. Work with CPU architects, software teams, and systems engineers to translate modeling results into actionable architectural recommendations.

This is a hybrid role based out of Santa Clara, CA or Austin, TX, contributing to Tenstorrent's CPU and system architecture efforts.

Experience Level

Senior-level. The posting notes candidates typically have 1–5 years of industry or equivalent research experience; Tenstorrent may assess and place candidates at the appropriate level during interview.

Responsibilities

Primary responsibilities focus on performance modeling, analysis, and collaboration to drive CPU architecture decisions.

• Develop and run cycle-accurate and analytic performance models and simulations for CPU cores and mid-core components.
• Use and extend simulation frameworks (e.g., gem5, SST, SimpleScalar, or equivalents) to evaluate microarchitecture changes.
• Analyze AI and HPC workload performance and identify bottlenecks across pipelines, caches, memory hierarchy, and interconnects.
• Translate modeling results into clear recommendations for architectural tradeoffs and design changes.
• Implement analysis tools, benchmarks, and scripts in C++, Python, or similar languages to automate experiments and data collection.
• Collaborate with CPU architects, hardware designers, and software engineers to validate designs and iterate on performance improvements.

Requirements

Technical must-haves and useful extras for successful candidates.

• Must-have: Experience with CPU core, mid-core, or microarchitecture development via industry work, research, or advanced academic projects.
• Must-have: Hands-on experience with performance modeling or simulation tools such as gem5, SST, SimpleScalar, or equivalent environments.
• Must-have: Strong programming skills for modeling and analysis (C++, Python, or similar).
• Must-have: Solid understanding of processor subsystems: instruction scheduling, branch prediction, caches, memory hierarchy, prefetching, and execution pipelines.
• Must-have: Eligibility to access U.S. export-controlled technology; employment may be contingent on citizenship, permanent residency, or ability to obtain required export licenses.
• Nice-to-have: Prior experience with RISC-V microarchitecture projects, AI/HPC workload characterization, or hardware-software co-design at scale.

Education Requirements

PhD preferred; MS considered. Preferred fields: Computer Engineering, Electrical Engineering, Computer Science, or a related technical field. Equivalent research or practical experience is acceptable.

About the Company

Company: Tenstorrent

Headquarters: Austin, Texas, United States

Tenstorrent is a technology company focused on designing innovative computing solutions. They are known for their expertise in the development of advanced hardware, including ASICs and SoCs, aimed at enhancing performance and efficiency in various applications.