Performance Architect (AI HW)

TL;DR

Performance Architect (AI HW): Modeling, analyzing, and optimizing AI workloads on the Tensix architecture with an accent on hardware-software co-design and PPA studies. Focus on developing performance models and simulators to drive feature evaluation and design optimization for next-generation AI systems.

Location: Hybrid (Toronto, ON; Austin, TX) or Remote

Compensation: $100k - $500k (including base and variable targets)

Company

hirify.global is developing high-performance AI hardware and a RISC-V CPU to redefine computing paradigms and efficiency.

What you will do

• Benchmark and analyze complex AI workloads across single and multi-node hardware configurations to guide architectural decisions.
• Develop and maintain performance models, simulators, and micro-benchmark suites for feature evaluation.
• Conduct detailed PPA (Performance, Power, Area) studies to assess design tradeoffs.
• Collaborate with RTL, Compiler, and Runtime teams to correlate performance models with actual silicon results.
• Translate deep learning workloads into architectural insights to ensure measurable performance gains.

Requirements

• Deep analytical skills and strong intuition for AI workload behavior and system-level performance bottlenecks.
• Proficiency in C++ and Python for simulation, modeling, and performance analysis.
• Experience working with heterogeneous compute systems.
• Ability to bridge the gap between software workloads and hardware architectural tradeoffs.
• Must be eligible to access U.S. export-controlled technology (EAR compliance), which may depend on citizenship or residency status.

Culture & Benefits

• Competitive compensation package and benefits.
• Collaborative environment focused on solving hard problems in AI hardware.
• Opportunity to work on cutting-edge RISC-V and AI compute fabric technology.
• Flexible work arrangements including hybrid and remote options.