The GM AV Organization's AI Solutions team needs to deploy machine learning models for inference from frameworks to autonomous vehicle hardware and develop the necessary ML runtime software.
Requirements
- Experience developing in a Linux environment
- Strong system fundamentals and coding abilities (C++, Python)
- Experience in ML model optimization, deployment, runtimes, or serving infrastructure
- Background in computer architecture and operating systems
- Experience with optimizing and/or evaluating complex software systems on dedicated hardware
- Understanding of how to design high-performant software components
- Experience with NVIDIA GPUs and CUDA
Responsibilities
- Deploy machine learning (ML) models that drive self-driving vehicles by leveraging ML compilers, hardware-aware ML optimizations, and ML runtimes, targeting both onboard (vehicle) and offboard (cloud/simulation) environments.
- Develop low-level ML runtime software and APIs to efficiently serve and execute models in onboard and offboard environments.
- Collaborate with engineers from Embodied AI, Kernels, Compilers, Architecture, and System Performance teams.
Other
- 3+ years of relevant industry experience
- BS, MS or PhD in CS, or related technical field
- Experience with low latency or real time systems
- Familiarity with ML hardware and performance of inference on ML hardware
- Experience with lower-levels of an accelerator software stack (i.e. drivers, runtimes, and user-level API) and the inter-layer interactions
