AI Hardware Performance Architect (Hybrid/Remote)

Tenstorrent Inc.; Toronto; Hybrid; CAD 139,000 - 694,000; Full time

Job summary

Tenstorrent Inc. seeks an AI Performance Architect to model, analyze, and optimize AI workloads on the Tensix architecture. You will shape hardware features and ensure design decisions yield measurable performance gains, connecting architecture, software, and RTL across next-gen AI systems.

The role is hybrid, based in Toronto, Austin, or remote, with offers aligned to experience level and market norms. Collaborative, data-driven, and curious engineers are encouraged to apply.

Qualifications

• Analytical engineer with strong intuition for AI workload behavior.
• Experience in C++ and Python for simulation and analysis.
• Ability to bridge software and hardware teams.
• Curious, data-driven, and focused on efficiency.

Responsibilities

• Benchmark and analyze AI workloads across single and multi-node configurations.
• Develop performance models, simulators, and micro-benchmark suites.
• Conduct PPA studies to inform hardware-software co-design.
• Collaborate with RTL, Compiler, and Runtime teams to map models to silicon.

Skills

AI workload analysis

C++

Python

Performance modeling

System-level design

RTL collaboration

Job description

Tenstorrent Inc. seeks an AI Performance Architect to model, analyze, and optimize AI workloads on the Tensix architecture. You will shape hardware features and ensure design decisions yield measurable performance gains, connecting architecture, software, and RTL across next-gen AI systems.

The role is hybrid, based in Toronto, Austin, or remote, with offers aligned to experience level and market norms. Collaborative, data-driven, and curious engineers are encouraged to apply.