Software Product Manager – AI Compute

Job Description

Intel's Data Center Group is hiring a Software Product Manager to own the product strategy, roadmap, and execution for AI compute middleware - the performance-critical software layers that enable AI training and inference on modern accelerators and CPUs.

Key Responsibilities:

• Define and own the product vision and roadmap for AI compute middleware
• Translate hardware capabilities into middleware features that unlock performance and usability
• Drive performance, scalability, and correctness requirements across AI training and inference workloads
• Partner with Framework PMs (PyTorch, JAX, vLLM), Driver/firmware teams, and Compiler/kernel engineers
• Prioritize investments based on model trends, customer requirements, and competitive benchmarks
• Ensure Day-0 enablement for new hardware, models, and software releases
• Align backlog and milestones across multiple product lines in a matrixed org
• Represent middleware strategy in internal reviews and external technical engagements

Requirements:

• Bachelor's + 6 years of experience, or Master's in CS, Software Engineering, MBA, or related field + 4 years
• 6+ years of software product management experience
• 2+ years of background in systems engineering or performance-critical software development
• Experience with AI training and inference workloads
• Familiarity with compute kernels, math libraries, and runtime performance tuning
• Knowledge of accelerator architectures (GPU/AI accelerators) and CPUs

Preferred:

• Hands-on experience with OneDNN, SYCL, Triton, CUDA-like, or kernel-level programming models
• Experience with AI frameworks or compiler stacks
• Experience with benchmarking, profiling, and performance analysis tools

This listing is from the company career pages — for more details, follow the link from the Apply button.