🛠 Tech Stack

💼 About This Role

You'll design and optimize high-performance ML kernels for large-scale language model training at Thinking Machines Lab. You'll collaborate with researchers to bridge algorithmic design with hardware efficiency and prototype kernel implementations across GPU architectures. This role offers the chance to work on cutting-edge AI infrastructure used by products like ChatGPT and PyTorch.

🎯 What You'll Do

• Design and implement custom ML kernels for core LLM operations.
• Optimize compute primitives to reduce memory bandwidth bottlenecks.
• Collaborate with research teams on kernel-level optimizations.
• Develop and maintain a library of reusable kernels and benchmarks.
• Contribute to infrastructure stability and scalability.

📋 Requirements

• Bachelor's degree in CS, EE, ML, physics, robotics, or similar.
• Proficiency in CUDA, CuTe, or Triton.
• Strong engineering skills with deep learning frameworks (PyTorch, JAX).
• Ability to analyze, profile, and optimize compute-intensive workloads.

✨ Nice to Have

• Experience training large-scale language models with tens of billions of parameters.
• Familiarity with tensor parallelism or pipeline parallelism.
• Experience with low-precision formats (FP8, INT8) or compiler stacks (XLA, TVM).

🎁 Benefits & Perks

• 🏖️ Unlimited PTO
• 🏥 Generous health, dental, and vision benefits
• 👶 Paid parental leave
• 📦 Relocation support

📨 Hiring Process

This is an evergreen role; applications are reviewed continuously and you may reapply every 6 months.

🚩 Heads Up

• Evergreen role may not have immediate opening.
• High salary range could imply intense work pressure.