Software Engineer - Machine Learning Systems

Summary

Join the MTIA (Meta Training & Inference Accelerator) Software team, part of the leading PyTorch AI framework organization. In this exciting role, you will contribute to a robust AI Compiler strategy designed to train and serve cutting-edge DL/ML model architectures, while ensuring high performance in production environments across diverse hardware platforms. This is your chance to engage with the core elements of the PyTorch framework, AI compilers, high-performance kernels, and tools that enhance machine learning workflows on both current and next-generation MTIA AI hardware. Collaborating with AI researchers, you will analyze deep learning models to optimize their performance on MTIA hardware and partner with hardware design teams for compiler innovations that drive efficiency in AI. Your expertise will also guide software development best practices in designing features and performance tuning, giving you invaluable experience in the development of machine learning compiler frameworks and participating in groundbreaking hardware-software co-design projects for AI applications.

Key Responsibilities

• Develop the software stack focusing on AI frameworks, compiler stack, and high-performance kernel development tailored for next-generation hardware architectures.
• Contribute to the PyTorch AI framework's core compilers to enhance the performance of state-of-the-art inference and training AI hardware accelerators.
• Analyze deep learning networks and implement innovative compiler optimization algorithms.
• Collaborate with AI research scientists on advancing deep learning models in fields including recommendation systems, generative AI, computer vision, and NLP.
• Optimize and fine-tune the performance of deep learning frameworks and software components.

Minimum Qualifications

• Strong programming skills in C/C++.
• Currently pursuing or recently obtained a Bachelor's degree in Computer Science, Computer Engineering, or a related technical field, to be completed prior to joining Meta.
• Experience in AI framework development or deep learning model acceleration on various hardware architectures.

Preferred Qualifications

• Bachelor's degree in Computer Science or related field with 4+ years of experience in AI framework development, or a Master's degree and 2+ years of experience, or a PhD in a relevant field.
• Familiarity with GPU, CPU, or AI hardware accelerator architectures.
• Experience working with frameworks such as PyTorch, Caffe2, TensorFlow, ONNX, or TensorRT.
• Background in high-performance kernel programming, CUDA, OpenMP, OpenCL, or relevant libraries like cuBLAS, cuDNN.
• Experience with compiler optimization techniques including loop optimization and vectorization, as well as familiarity with tools like MLIR, LLVM, or TVM.
• Proven ability to integrate AI tools to optimize workflows and achieve measurable improvements in efficiency and quality.
• Understanding of responsible AI practices including bias mitigation and quality assessment.
• Continuous development of AI skills and knowledge of emerging technologies.

Public Compensation

$58.65/hour to $181,000/year + bonus + equity + benefits

Industry: Internet

Equal Opportunity

Meta is an Equal Employment Opportunity and Affirmative Action employer, committed to a diverse workplace free of discrimination based on legally protected characteristics. We also consider qualified applicants with criminal histories per applicable law. Additionally, Meta participates in the E-Verify program where required.

Meta is dedicated to ensuring an inclusive recruiting process and provides reasonable accommodations for candidates with disabilities. If you require any assistance or accommodations due to a disability, please inform us.