AI Infra Engineer – GPU & AI Accelerators

Oracle

United States

On-site

USD 97,000 - 306,000

Full time

5 days ago

Be an early applicant

Get a reply from this employer

Your resume and cover letter, tailored to exactly what they’re hiring for.

Get past ATS filters

Job summary

Oracle Hardware Platform Development Engineering seeks an AI Systems Engineer to bring up new GPU/AI accelerator platforms for OCI, run workloads, and analyze performance in real hardware conditions. You will compare architectures, diagnose hardware-software integration issues, and craft actionable recommendations for AI training and inference workloads.

Ideal candidates have 8+ years in large-scale AI systems, strong debugging skills, and hands-on experience with CUDA/ROCm stacks, high-speed

Qualifications

Strong knowledge of AI/GPU or AI/CPU platform architectures.
Experience with multi-architecture server platforms and vendors (x86/ARM).
Excellent written and verbal communication across engineering teams and executives.
Familiarity with high-speed buses/interconnects in compute/AI platforms.
Ability to debug complex hardware-software interactions and resolve issues.
Experience profiling training workloads and runtime to optimize performance.

Responsibilities

Evaluate and characterize next-gen GPU/AI accelerator platforms for Oracle Cloud Infrastructure.
Bring up hardware platforms and enable AI training/inference software stacks.
Run representative workloads and analyze system performance under real operating conditions.
Identify workload bounds (HBM/memory, compute, scale-up/scale-out) and report findings.
Perform comparative architecture analysis across GPUs and accelerators and translate findings into recommendations.
Collaborate with internal teams and partners to shape Oracle’s AI infrastructure.