Senior AI Algorithm Engineer in oneDNN

Responsibilities

• Design, develop, and optimize new features and algorithms for oneDNN targeting Intel processors, Intel Processor Graphics, and Intel discrete GPUs.
• Perform performance analysis and optimization to achieve best-in-class deep-learning inference and training throughput on current and next-generation Intel platforms.
• Develop hardware-specific parallel algorithms, including multithreading, vectorization, and memory-layout optimizations.
• Contribute to assembly-level programming and low-level performance tuning for Intel microarchitectures.
• Collaborate with cross-functional teams across software engineering, architecture, and AI performance to ensure strong integration with Intel’s broader AI ecosystem.
• Engage with the open-source community, participate in code reviews, and maintain high-quality coding and documentation standards.

Qualifications

• BSc, MSc or PhD in Mathematics, Physics, Computer Science or in a related field
• 5+ years of experience in the following areas:
• C++
• Algorithms and data structures, or Mathematical background
• Low-level Performance Optimizations, preferably on GPUs
• Preferred Qualifications:
• 3 years+ High-performance computing (HPC) applications development
• 1 year+ Machine learning and deep learning algorithms
• 1 year+ Agile software development environment
• 1 year+ Intel development tools
• Software libraries design and architecture
• Background in Linear algebra solvers, matrix-vector operations, or Fast Fourier Transforms
• Software development on Linux
• GPU optimizations (OpenCL, CUDA, SYCL/DPC++, C for Metal or similar)
• Parallel programming (OpenMP, TBB, or MPI)

Benefits

Intel offers a total compensation package that ranks among the best in the industry. It consists of competitive pay, stock bonuses, and benefit programs which include health, retirement, and vacation.