Tenstorrent is a leader in cutting-edge AI technology, aiming to revolutionize performance, ease of use, and cost efficiency. As AI redefines computing, Tenstorrent is unifying innovations across software models, compilers, platforms, networking, and semiconductors. Our team has developed a high-performance RISC-V CPU from scratch and is passionate about building the best AI platform.
As a Machine Learning Engineer on the AI Models team at Tenstorrent, you will bring up and optimize cutting-edge AI models for our custom AI devices. You will experiment, optimize, and push boundaries to solve real-world problems. This role is ideal for individuals who are passionate about ML and want to work on models used in real-world applications.
This is a hybrid role based in Warsaw or Gdansk, Poland, with consideration for remote candidates on a case-by-case basis. We welcome candidates of all experience levels, and the appropriate level will be determined during the interview process.
Who You Are
- Proficient in Python programming and experienced with PyTorch for deep learning model development.
- Curious, experimental, and driven to understand and improve complex systems.
- Possess a deep understanding of ML model architectures for optimizing components and their interactions.
- Collaborative and eager to work with software and hardware teams.
What We Need
- Experience bringing state-of-the-art ML models to new hardware platforms.
- Ability to analyze performance bottlenecks, fine-tune architectures, and enhance model accuracy and robustness.
- Familiarity with model optimization techniques (e.g., quantization, sparsity, attention) and hardware features like matrix engines, caches, and memory hierarchies.
- Strong understanding of current ML research and the ability to apply it to practical improvements.
What You Will Learn
- How to deploy real ML models on custom AI accelerators.
- Techniques for optimizing ML model performance from application to silicon level.
- The process of moving ML models from research papers to production-ready deployments.
- How to collaborate with compiler, kernel, and hardware teams to drive new features, performance optimizations, and fixes.
Tenstorrent offers a highly competitive compensation package and benefits.