Software Engineer, TT-Distributed

About the role

Tenstorrent is at the forefront of AI innovation, aiming to redefine performance, usability, and cost-effectiveness in AI technology. We are building a comprehensive AI platform that integrates software models, compilers, hardware, and networking. Join our team to contribute to the development of high-performance distributed systems for AI and High-Performance Computing.

Key facts

What you'll do

• Design, build, and refine distributed software systems that manage computation and communication across clusters of AI processors and CPUs.
• Create interfaces for distributed programming that support data and tensor parallelism in AI workloads.
• Utilize MPI and related technologies to scale programming models across numerous machines and compute units.
• Develop reliable and high-performance systems using inter-process communication, network sockets, and distributed communication methods.
• Construct and maintain tools for testing, debugging, profiling, and monitoring large-scale distributed tasks, and work with model and systems teams during cluster setup.

Requirements

• Proficient in C or C++ with a strong understanding of systems programming, operating systems, and distributed systems concepts.
• Experience with distributed computing, including inter-process communication, socket programming, and cluster resource management.
• Ability to analyze scalability, fault tolerance, and performance in multi-node environments.
• A curious mindset and a willingness to question traditional approaches to distributed system design.
• A drive to become a technical expert in large-scale distributed AI infrastructure.
• Minimum 2 years of relevant professional experience.

Nice to have

• Experience with Python for scripting and tooling.
• Familiarity with AI frameworks like PyTorch or TensorFlow.
• Knowledge of RISC-V architecture.

Skills & tools

• C/C++
• Distributed Systems
• Operating Systems
• MPI
• IPC
• Socket Programming
• Cluster Computing

Practical notes

• This position is hybrid and based in Toronto, Ontario, Canada.
• Compensation is between $100,000 and $500,000, varying based on experience, skills, background, and location.
• Employment is contingent upon eligibility to access U.S. export-controlled technology, as per U.S. export regulations. This may require verification of citizenship/permanent residency status or obtaining a license.
• Tenstorrent offers competitive compensation and benefits.
• We are an equal opportunity employer.