Tesla Sr. Software Engineer PyTorch Export AI Inference 

United States, California, Palo Alto 

555535599

Our AI Inference team puts ML models into production - we train and deploy large neural networks for efficient inference on compute-constrained edge devices (CPU / GPU / AI ASIC). The nature of this role is multi-disciplinary - you will work at the intersection of machine learning and systems by building the ML frameworks and infrastructure that enable the seamless deployment, and inference of all neural networks that run on Autopilot and Optimus.

• Build robust AI frameworks to lower neural networks to edge devices. Knowledge of torch export/FX transforms for the frontend, or LLVM-based compilers for the backend is preferred here
• Deploy state-of-the-art neural networks on low-precision heterogenous compute, including Tesla’s in-house AI ASIC, with an aim to maximize network performance while minimizing latency
• Collaborate with AI scientists and compiler engineers to effectively run large models in low precision. The ideal candidate is familiar with recent advancements in network architectures and inference optimizations for the current state of the art models
• Design and implement custom GPU kernels (CUDA / OpenCL) for efficient training and processing of network outputs

• Proficiency with Python
• Proficiency with PyTorch (PyTorch Export) or JAX
• Experience with ML Optimization, ML Architecture or Model Integration
• Proficiency with training and deploying neural networks for real-world AI
• Proficiency with computer systems and computer architecture
• Experience with deployment, compiler, runtime
• Experience with CUDA