Intel
AI Performance Library Architect
$170K–$315K
US, Oregon, Hillsboro +1 · Full-time · Posted Sep 17, 2026
- Location
- US, Oregon, Hillsboro
US, California, Folsom - Experience
- 5+ years across the stated areas; preferred experience includes 3 years+ and 1 year+ areas as listed.
- Education
- Master’s degree or Ph.D. degree in Mathematics, Physics, Computer Science, or a relevant STEM field.
About the job
Designs, develops, and maintains oneDNN functionality to improve performance for AI workloads on Intel CPUs and GPUs, while supporting AI framework developers and open-source contributors.
What you'll do
- Design, develop, and maintain performance-critical oneDNN functionality.
- Support developers optimizing AI frameworks and workloads for Intel CPUs and GPUs.
- Collaborate with the cross-platform AI software community contributing to oneDNN.
What they ask for
- Master’s or Ph.D. degree in Mathematics, Physics, Computer Science, or a relevant STEM field.
- 5+ years of experience across the listed areas: C and C++, open-source software, library design and architecture, linear algebra algorithms, performance optimization, floating-point arithmetic and numerical stability, Linux software development, and low-level performance optimization.
Nice to have
- 3 years+ in machine learning/deep learning algorithms or HPC application development; 3 year+ implementing floating-point transcendental functions.
- 1 year+ with algorithms for non-IEEE low-precision data types; 1 year+ AI-assisted software development.
Pay
$170,500.00-315,490.00 USD (target compensation; base salary or total pay not specified)
Tools
oneDNN, C, C++, BLAS, LAPACK, PyTorch, CUDA, OpenCL
AI focus
Develops and optimizes a neural-network performance library used by AI frameworks and workloads; preferred experience includes machine learning, deep learning, and AI-assisted software development.
Model-written summary · the original posting is authoritative