Principal Engineer - AI Networking

Oracle; United States; On-site; USD 99,600 - 234,600; Full time

Benefits offered by this job

Potential bonus

Equity options

Compensation deferral

Job summary

Oracle is looking for an RDMA software engineer to design, implement, and optimize networking infrastructure for AI workloads. Candidates should have extensive experience in high-performance networking, systems programming, and RDMA technologies such as RoCEv2 and InfiniBand.

The role emphasizes collaboration with technical leaders, requiring strong problem-solving skills and the ability to enhance software components to support advanced AI systems.

Qualifications

• 7+ years of software engineering experience in systems software, networking, or distributed systems.
• Strong hands-on expertise with RDMA technologies.
• Proficiency in debugging and optimizing performance-critical software.

Responsibilities

• Design and optimize RDMA-based software components for AI infrastructure.
• Collaborate with teams to deliver high-performance solutions.
• Develop performance analysis tools for distributed systems.

Skills

Software engineering fundamentals

Systems programming

Performance optimization

Collaboration skills

Ownership mindset

AI systems infrastructure building

Education

Bachelor's degree in Computer Science or related field

Advanced degree preferred

Tools

C/C++

Linux systems programming

NCCL

MPI

Kubernetes

Job description

Job Description

The ideal candidate is an experienced RDMA software engineer with a strong background in high-performance networking, distributed communication systems, and systems programming. You will work closely with senior technical leaders to design, implement, optimize, and operate critical networking infrastructure used by large-scale AI training and inference workloads.

This is a hands‑on engineering role requiring deep technical expertise, strong software development skills, and a passion for solving complex performance and scalability challenges.

What You'll Bring

• Strong software engineering fundamentals and systems programming expertise.
• Deep interest in RDMA, high-performance networking, and distributed communication systems.
• Ability to diagnose and solve complex performance and scalability problems.
• Strong collaboration and communication skills in cross‑functional engineering environments.
• Ownership mindset with the ability to independently drive technical initiatives from design through production deployment.
• Passion for building infrastructure that enables next‑generation AI systems.

Key Responsibilities

• Design, develop, and optimize RDMA‑based software components and services for large‑scale AI infrastructure.
• Build and enhance collective communication frameworks, transport layers, and communication libraries used by distributed AI workloads.
• Develop congestion management, load balancing, resiliency, and failover capabilities for RDMA‑based networks.
• Analyze and improve communication performance across networking, GPU, and software stacks.
• Design and implement scalable distributed systems supporting AI training and inference environments.
• Collaborate with networking, AI infrastructure, hardware, and cloud platform teams to deliver high‑performance solutions.
• Investigate and resolve complex networking, performance, and reliability issues in production environments.
• Develop observability, telemetry, debugging, and performance analysis tools for distributed communication systems.
• Contribute to architectural design discussions and technical direction for networking platforms.
• Participate in code reviews and help maintain engineering excellence across the team.

Minimum Qualifications

• Bachelor's degree in Computer Science, Computer Engineering, Electrical Engineering, or related field; advanced degree preferred.
• 7+ years of software engineering experience in systems software, networking, distributed systems, or infrastructure platforms.
• Strong hands‑on expertise with RDMA technologies, including RoCEv2 and/or InfiniBand.
• Experience developing RDMA‑enabled software, communication libraries, networking services, or distributed infrastructure.
• Strong understanding of RDMA programming concepts, including queue pairs, completion queues, memory registration, verbs, and transport semantics.
• Proficiency in C/C++ and Linux systems programming.
• Experience debugging and optimizing performance‑critical software systems.
• Solid understanding of networking fundamentals, operating systems, and distributed systems concepts.

Preferred Qualifications

• Experience with collective communication frameworks and libraries such as NCCL, RCCL, MPI, UCX, UCC, XCCL, or similar technologies.
• Experience supporting AI/ML infrastructure and distributed training environments.
• Knowledge of GPUDirect RDMA and GPU‑aware communication technologies.
• Experience developing congestion management, traffic engineering, or network resiliency solutions.
• Familiarity with large‑scale GPU clusters and high-performance computing environments.
• Experience building services and infrastructure operating directly over RDMA transports.
• Familiarity with distributed training frameworks such as PyTorch, DeepSpeed, Megatron‑LM, TensorFlow, or JAX.
• Experience with Kubernetes, containers, and cloud infrastructure platforms.
• Understanding of performance profiling and benchmarking tools for networking and distributed systems.

Compensation and Benefits

US: Hiring Range in USD from $99,600 – $234,600 per year, with potential bonus, equity, and compensation deferral.

Disclaimers

Certain U.S. based or U.S. customer or client‑facing roles may be required to comply with applicable requirements, such as immunization/occupational health mandates, and/or drug testing requirements.

EEO Statement

Oracle is an Equal Employment Opportunity Employer. All qualified applicants will receive consideration for employment without regard to race, color, religion, sex, national origin, sexual orientation, gender identity, disability and protected veterans’ status, or any other characteristic protected by law.