Software Engineer, Metal Runtime (Core Systems)

About Tenstorrent

Tenstorrent is at the forefront of AI innovation, reshaping expectations regarding performance, usability, and cost-efficiency. As AI transforms computing paradigms, integrated development across software models, compilers, platforms, networking, and semiconductor technologies is essential. Our diverse team has engineered a high-performance RISC-V CPU from the ground up, driven by a passion for AI and commitment to creating the optimal AI platform. We embrace collaboration, curiosity, and tackling challenging problems.

Role Overview

As a Software Engineer on the Metal Runtime team, you will develop the low-level software that controls our AI accelerators. Your focus will be creating and refining high-performance runtime systems that operate directly on hardware, emphasizing scheduling, memory management, and efficient execution on massively parallel processors. Runtime systems are critical to performance, managing hardware resource utilization, data flow, and scalable workload execution.

Key Responsibilities

• Build and enhance the Metal runtime executing on AI acceleration hardware.
• Design bare-metal systems governing compute execution, scheduling, and resource allocation.
• Optimize memory movement and data pathways across parallel hardware architectures.
• Partner closely with hardware engineers to boost performance and efficiency.
• Analyze, debug, and fine-tune runtime operations within complex distributed environments.

Candidate Profile

• Proficient in C or C++ with a strong affinity for hardware-level programming.
• Adept in managing threads, processors, and memory movement; skilled in writing parallel, high-performance code.
• Enjoy optimizing scheduling, execution, and memory systems for scalability and speed.
• Experienced in debugging at the intersection of hardware and software, with holistic system understanding.
• Enthusiastic about the integration of runtimes and bare-metal software to power AI applications.
• Collaborative mindset, thriving in tackling complex technical challenges with a skilled team.

Learning Opportunities

• Gain expertise in designing advanced runtime systems for emerging AI hardware.
• Understand interactions between execution, scheduling, and memory at scale.
• Experience optimization of custom ASICs and co-processors across silicon.
• Develop advanced debugging skills for performance-critical environments.
• Witness software and hardware co-evolution in a dynamic AI startup setting.

Compensation and Legal Considerations

Compensation ranges between $100,000 and $500,000 annually, combining base salary and variable components, depending on experience, skills, education, background, and location.

The role requires eligibility to access US export-controlled technologies. Compliance with US Export Administration Regulations is mandatory, potentially affecting offer validity based on citizenship, permanent residency, or licensing approvals. Offers may be rescinded if compliance cannot be ensured.

Location and Work Model

This position is based onsite in the Greater Toronto Area, Canada, with hybrid opportunities available at Santa Clara, CA, and Austin, TX.