$170,500 – $315,490
Listed on Intel’s own careers site. You apply with them directly — we never stand between you and the employer.
What this role is
Intel is seeking a software engineer to design and optimize neural network performance libraries in the oneDNN project, supporting AI frameworks across CPUs and GPUs. The role suits developers with deep expertise in performance engineering, low-level optimization, and linear algebra who want to influence AI infrastructure used across the industry.
Our summary, not Intel’s wording. The full posting is on their site.
What they ask for
Required
- Master's or Ph.D. in Mathematics, Physics, Computer Science, or relevant STEM field
- 5+ years C and C++
- 5+ years maintaining or contributing to open-source software projects
- 5+ years software libraries design and architecture
- 5+ years implementing linear algebra algorithms from BLAS, LAPACK, or PyTorch
- 5+ years performance engineering and software performance optimization
- 5+ years working with floating point arithmetic and numerical stability
- 5+ years software development on Linux
- 5+ years low-level performance optimization using CUDA, x86 assembly, intrinsics, or OpenCL
Nice to have
- 3+ years machine learning and deep learning algorithms or HPC development
- 3+ years floating point implementations of transcendental functions
- 1+ years algorithms for non-IEEE low precision data types like bfloat16, fp8, fp4
- 1+ years AI-assisted software development