Design Verification Lead, AI Hardware

Job Title

Design Verification Lead, AI Hardware

Role Summary

Lead a verification team for next-generation AI SoC hardware, defining verification strategy and ensuring functional and performance correctness of AI cores and on-chip networks. The role is hybrid, based out of Toronto, ON, and requires close collaboration with architecture, software, and silicon teams.

Experience Level

Senior-level (Lead). The posting targets a verification lead with significant ASIC/SoC and tape-out experience; specific years of experience were not stated.

Responsibilities

Drive verification strategy and execution across the project lifecycle and mentor the verification team.

• Define and lead verification plans for AI cores, focusing on functionality and performance under real workloads.
• Design coverage-driven test strategies for mixed-precision datatypes (FP8, BF16, INT8) and tensor operations.
• Validate and benchmark on-chip network (NoC) performance, throughput, bandwidth, and congestion scenarios.
• Implement resiliency testing and error-injection scenarios for concurrency and multi-core synchronization edge cases.
• Develop portable, platform-agnostic tests that scale from subsystem simulation to FPGA bring-up and silicon validation.
• Collaborate with architects and software teams to align verification with performance targets and workload behavior.
• Provide technical leadership and mentoring to elevate team practices and automation frameworks.

Requirements

Core must-have skills and technologies for immediate contribution, followed by useful but non-mandatory skills.

• Must-have: Proven leadership of ASIC/SoC verification teams and track record through complex tape-outs.
• Must-have: Deep experience with verification methodologies and languages (UVM, SystemVerilog, cocotb).
• Must-have: Expertise in AI/ML compute patterns and mixed-precision verification (FP8, BF16, INT8).
• Must-have: Strong experience validating NoC behavior under high-demand traffic and performance stress.
• Must-have: Experience implementing error-injection, concurrency controls, and multi-core synchronization tests.
• Must-have: Ability to develop portable verification strategies that scale from simulation to FPGA and silicon.
• Nice-to-have: FPGA/silicon bring-up experience, scripting and automation (e.g. Python), and familiarity with AI workload characteristics.

Education Requirements

Not specified.

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.