Lead Machine Learning Compiler Engineer

General Motors

United States

United States

Postuler Maintenant

À propos

Join Our Mission At our core, we aim for Zero Crashes, Zero Emissions, and Zero Congestion in the realm of autonomous and assisted driving. Our commitment to innovation drives us in developing fully self-driving systems that contribute to safer and more sustainable mobility. The AI Kernels & Compilers team plays a pivotal role in this mission, turning groundbreaking research in perception, prediction, and planning into high-quality, scalable software that performs reliably in real-world vehicles. We are continuously improving methodologies for model export, kernel development, and performance engineering, focusing on enhancing situational awareness and vehicle behavior. If you are passionate about advancing compiler and kernel technologies that make a tangible impact in the world of autonomous vehicles, we have an exciting opportunity for you. About the Team Our AI Compiler team is integral to integrating advanced AI models into vehicles. We optimize and transform high-level models for efficient inference on GPUs to power our cutting-edge features in autonomous and assisted driving. We focus on graph lowering, operator coverage, kernel integration, and deployment tools to maximize performance while ensuring reliability. Collaboration across AI Deployments, AI Solutions, Runtime, and AI Kernels teams is vital for delivering research ideas into production efficiently. As part of a talented group of compiler, systems, and GPU engineers, you’ll face complex challenges and explore MLIR/ONNX and CUDA/TensorRT intricacies. We value sound reasoning and fundamental engineering principles, creating an environment for excellence in shaping the future of automated driving. Your Role As a Lead Compiler Engineer in the AI Kernels & Compilers team, you will drive the enhancement of the compilation stack that transforms high-level models into optimized inference artifacts for our autonomous and assisted driving platforms. Your contributions will help establish tools and pipelines that streamline the model compilation process for ML engineers organization-wide. Your responsibilities will encompass advancing a cutting-edge model export and compilation pipeline, ensuring efficient transitions from high-level model graphs to specific inference engines while maintaining a balance of compilation speed, fidelity, and on-vehicle latency. You will also develop tools for numerical accuracy, performance regression identification, and diagnostics for model authors. If you are ready to work at the intersection of
compilers, performance engineering, and real-world autonomy , this position will place you at the forefront of innovation within the vehicle ecosystem. Your Responsibilities Enhance the model compilation toolchain
to efficiently deploy large-scale models for perception, prediction, and planning. Design and implement new compiler passes and analyses
to improve build times, memory efficiency, and runtime performance, while ensuring compliance with safety and reliability standards. Collaborate collaboratively
with kernels, runtime, and hardware teams to develop interfaces and enhance accelerator capabilities for optimal platform performance. Establish best practices
for model export, validation, and debugging to support rapid, iterative processes for autonomous vehicle teams with reproducible performance metrics. Your Skills & Abilities (Required Qualifications) At least 3 years of experience in compiler development. Proficient in ML frameworks (e.g., PyTorch, TensorFlow, or JAX) and the software stack (including ONNX, MLIR, XLA, TVM, TensorRT). Expertise in writing production-quality Python/C++ code. Sound understanding of the software development lifecycle: coding, debugging, optimization, testing, and integration. Bachelor’s degree or higher in Computer Science, Computer Engineering, Electrical Engineering, or a related field. Preferred Qualifications (Competitive Edge) Experience in building and optimizing ONNX-based model export and deployment pipelines. Knowledge of GPU programming (CUDA) and familiarity with the ML software stack (e.g., cuDNN, cuBLAS). Experience with ML accelerators and hardware architectures. Background in developing and deploying machine learning models. Compensation:
Salary range for this position is $128,700 to $261,300, dependent on experience and qualifications. Bonus Structure:
Includes competitive incentive pay based on performance metrics. Benefits:
Comprehensive health programs including medical, dental, and vision coverage, flexible spending accounts, retirement plans, tuition assistance, and GM vehicle discounts. About Us We are committed to leading initiatives that promote a world with Zero Crashes, Zero Emissions, and Zero Congestion, striving for a safer and more sustainable future for everyone. Why Join Us? We believe in driving meaningful change through our actions and culture, ensuring every employee feels valued as part of the General Motors team. Equal Opportunity Employment General Motors embraces diversity and is committed to an equitable workplace. We encourage applicants from all backgrounds who align with our mission and values to apply. Accommodations General Motors welcomes applicants with disabilities and will provide reasonable accommodations upon request during the application process. Join us in driving change for a better, safer, and more equitable world through our innovative actions.

United States

Compétences linguistiques

English

Avis aux utilisateurs

Cette offre provient d’une plateforme partenaire de TieTalent. Cliquez sur « Postuler maintenant » pour soumettre votre candidature directement sur leur site.

Postuler Maintenant