As a Software Engineer on the Acceleration Kernel Development team at Tenstorrent, you’ll work at the intersection of software and hardware performance. You’ll be writing low-level code that directly powers high-efficiency machine learning workloads, optimizing every cycle, every memory move, every instruction. If you're motivated by performance, precision, and real impact, this is where your skills will shine.

This role is hybrid, based out of Santa Clara, CA or remote.

We welcome candidates at various experience levels for this role. During the interview process, candidates will be assessed for the appropriate level, and offers will align with that level, which may differ from the one in this posting.

Who You Are

• A developer who loves high performance code, parallel algorithms, wrangling bits, optimizing compute, and making hardware fly.
• Great in C/C++ and able to build fast, efficient code from the ground up.
• Obsessed with performance and precision, especially in ML workloads.
• Motivated by complex problems and thrives in collaborative, fast-moving environments.

What We Need

• Expertise in building and optimizing compute kernels for parallel ML and high-performance workloads.
• Ability to analyze and tune instruction-level performance across latency, memory, and bandwidth.
• A collaborative mindset to work closely with ML engineers and integrate optimizations into production.
• Ownership of debugging, profiling, and maintaining a fast, reliable low-level software stack.

What You Will Learn

• The art of pushing AI hardware to its limits by shaping how kernels are written and executed.
• How to integrate kernel work into ML frameworks and real-world training pipelines.
• Skills in tuning performance on cutting-edge architectures with top-tier hardware engineers.
• Expertise in keeping code lean, reliable, and scalable even under heavy workloads.