Senior Software Engineer, AI Compiler
TenstorrentJob Title
Senior Software Engineer, AI Compiler
Role Summary
Work on TT-Forge, an MLIR-based compiler that maps machine-learning models to Tenstorrent hardware. The role focuses on compiler optimization, creating and maintaining MLIR dialects and transformation passes, and enabling performant execution across single- and multi-chip configurations.
This position is hybrid and based in Toronto, ON, Canada; you will collaborate with hardware and software engineers and contribute to open-source tooling.
Experience Level
Senior. The posting indicates candidates may be assessed at multiple levels during interview, but the advertised title is senior.
Responsibilities
Primary responsibilities include developing and maintaining compiler components and ensuring efficient execution of ML workloads on Tenstorrent hardware.
- Design and implement compiler transformation passes and optimizations in MLIR.
- Create and maintain custom MLIR dialects (e.g., TTIR, TTNN, TTKernel) and lowering paths.
- Bridge high-level frameworks (PyTorch, JAX, TensorFlow) to low-level execution targets.
- Work on performance tuning, multi-chip scaling, and generation of efficient kernels.
- Collaborate with hardware, firmware, and systems teams to validate and debug compiler output on silicon.
- Participate in open-source development and code reviews; improve developer tooling and usability.
- Mentor junior engineers and help shape engineering practices for the compiler stack.
Requirements
Must-have technical skills and expectations; additional desirable skills follow.
Must-have:
- Fluency in C++ and Python and experience building complex systems.
- Experience connecting ML frameworks to execution backends or similar systems-level engineering.
- Strong interest or experience with compiler optimization and MLIR-based toolchains.
- Familiarity with ML frameworks such as PyTorch, JAX, or TensorFlow, or deep systems programming experience.
- Ability to collaborate across hardware and software teams and contribute to open-source projects.
- Employment is contingent on eligibility to access U.S. export-controlled technology; offers may depend on citizenship/permanent residency or license approval.
Nice-to-have:
- Experience creating custom MLIR dialects or kernel libraries.
- Experience with multi-chip scaling, performance engineering, or compiler tuning tools.
- Prior work on machine-learning compiler stacks, code generation, or low-level runtime systems.
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.
