Senior AI Algorithm Engineer — Remote HPC Performance

Intel Corporation; Santa Clara (CA); On-site; USD 195,000 - 276,000; Full time

Benefits offered by this job

Stock bonuses

Health insurance

Retirement plan

Paid time off

Employee discounts

Meal card / cafeteria benefits

Job summary

Intel Corporation seeks a Software Development Engineer to advance oneDNN, a cross‑platform, open‑source library for DL. You will design and optimize features for Intel CPUs, Graphics, and discrete GPUs, perform performance tuning, and collaborate across teams to ensure strong integration with the AI ecosystem.

The role emphasizes low‑level optimization, parallel programming, and open‑source engagement, with a focus on high‑throughput inference and training across Intel hardware generations.

Qualifications

• BSc, MSc or PhD in Mathematics, Physics, Computer Science or related field.
• 5+ years in C++, algorithms/DS, and low-level performance optimizations (GPU preferred).

Responsibilities

• Design, develop, and optimize oneDNN features for Intel processors and GPUs.
• Perform performance analysis to boost inference/training throughput.
• Develop hardware-specific parallel algorithms and memory-layout optimizations.
• Contribute to assembly-level programming and low-level tuning for Intel microarchitectures.
• Collaborate with cross-functional teams to integrate with the AI ecosystem.

Skills

C++

Algorithms

Low-level optimizations

HPC apps development

Machine learning basics

Parallel programming

Education

BSc/MSc/PhD in Mathematics, Physics, CS or related field

Tools

OpenCL

CUDA

SYCL/DPC++

C for Metal or similar

Linux development

Job description

Intel Corporation seeks a Software Development Engineer to advance oneDNN, a cross‑platform, open‑source library for DL. You will design and optimize features for Intel CPUs, Graphics, and discrete GPUs, perform performance tuning, and collaborate across teams to ensure strong integration with the AI ecosystem.

The role emphasizes low‑level optimization, parallel programming, and open‑source engagement, with a focus on high‑throughput inference and training across Intel hardware generations.