Principal Developer, AI Networking

Description

The company has spent more than 25 years advancing computer graphics, PC gaming, and accelerated computing, and is now applying AI to shape the next phase of computing, where GPUs serve as the processing core for computers, robots, and self-driving cars that interpret their surroundings.

The AI Networking Codesign and Benchmarking R&D group is hiring a senior software engineer to profile, analyze, and tune AI workloads running on large GPU and CPU clusters that train and serve distributed LLMs.

The main focus is collective communication and networking. The work spans hardware such as HCAs, switches, CPUs, GPUs, and full systems, along with software layers including LLM applications, machine learning frameworks, and communication and computing libraries.

The team is growing quickly and wants engineers who work independently and bring creativity and a strong interest in technology.

Responsibilities

• Study AI workloads and deep learning models built for large-scale LLM training and inference on the company's supercomputers, with emphasis on distributed systems, high-performance networking, and the company communication libraries.
• Benchmark, profile, and analyze performance to locate bottlenecks, with particular attention to networking, and propose improvements.
• Build PyTorch trace-based tools for profiling, analysis, and replay that support benchmarking, debugging, and network co-design for LLM workloads.
• Work with hardware and software teams to supply performance analysis findings.
• Write performance test plans, set expectations for new technologies and solutions, and drive results toward performance targets.

Requirements

• Bachelor's degree in computer science or software engineering, or equivalent experience.
• 15+ years of work with high-performance networking, including RDMA, MPI, NCCL, and SHARP.
• Proven skill in performance evaluation methods and approaches.
• Hands-on experience with the company GPUs and CUDA, plus familiarity with deep learning frameworks such as TensorFlow or PyTorch.
• Working knowledge of collective communication libraries such as NCCL and of protocols such as RoCE and RDMA.
• Ability to learn quickly on your own, with strong analytical and problem-solving skills.
• Proficiency in Python, Bash, and C++.
• Experience working in container-based development environments.
• Clear communicator who collaborates well with others.
• Ways to stand out: extensive hands-on work with AI workloads and benchmarking of distributed LLM training.
• Ways to stand out: knowledge of PyTorch, CUDA, and NCCL libraries.
• Ways to stand out: broad understanding of systems, including Intel, AMD, or ARM CPUs, the company GPUs, HCAs, memory, and PCI.
• Ways to stand out: solid skill in performance evaluation using current tools.

Conditions

Base salary depends on location, experience, and other objective job-related factors. Level 6 ranges from 272,000 USD to 431,250 USD, and Level 7 ranges from 320,000 USD to 488,750 USD.

Applications will be accepted at least until November 3, 2026.

This posting is for an existing opening.

The company uses AI tools in parts of its recruiting process.

The company is an equal opportunity employer and does not discriminate on the basis of race, religion, color, national origin, gender, gender expression, sexual orientation, age, marital status, veteran status, disability status, or any other legally protected characteristic.

Benefits

• Competitive salary and a generous benefits package.