Senior Deep Learning Algorithm Engineer

Details

NVIDIA is hiring engineers for the core AI Frameworks team (Megatron Core and NeMo Framework) to design, develop, and optimize diverse real-world workloads. Megatron Core and NeMo are open-source, scalable, cloud-native frameworks for researchers and developers working on Large Language Models (LLM) and Multimodal (MM) foundation model pretraining and post-training. The team provides end-to-end model training, including pretraining, reasoning, alignment, customization, evaluation, deployment and tooling to optimize performance and user experience.

Responsibilities

• Expand Megatron Core and NeMo Framework capabilities so users can develop, train, and optimize models.
• Design and implement distributed training algorithms, model-parallel paradigms, and model optimizations.
• Define robust APIs and expand toolkits and libraries for coherence and completeness.
• Analyze and tune performance across the software stack; collaborate with internal partners and the open-source community to implement highly optimized solutions.
• Develop algorithms for AI/DL, data analytics, machine learning, or scientific computing.
• Contribute to and advance open-source projects such as NeMo-RL, Megatron Core, and the NeMo Framework.
• Solve large-scale, end-to-end AI training and inference challenges across the model lifecycle: orchestration, data pre-processing, training, tuning, and deployment.
• Work across computer architecture, libraries, frameworks, AI applications, and the full software stack.
• Research, prototype, and develop robust and scalable AI tools and pipelines.
• Perform performance tuning and optimizations, including model training and fine-tuning with mixed precision recipes on modern NVIDIA GPU architectures.

Requirements

• MS, PhD or equivalent experience in Computer Science, AI, Applied Math, or related fields.
• 5+ years of industry experience.
• Experience with AI frameworks (examples given: PyTorch, JAX, Ray) and/or inference/deployment environments (examples: TRTLLM, vLLM, SGLang).
• Proficient in Python programming, software design, debugging, performance analysis, test design and documentation.
• Strong understanding of AI / deep-learning fundamentals and practical applications.
• Proven record of working effectively across multiple engineering initiatives and improving AI libraries with new innovations.

Preferred / Ways to stand out

• Hands-on experience in large-scale AI training and deep understanding of core compute system concepts (latency/throughput bottlenecks, pipelining, multiprocessing) with demonstrated excellence in performance analysis and tuning.
• Prior experience with Reinforcement Learning algorithms and compute patterns.
• Expertise in distributed computing, model parallelism, and mixed precision training.
• Experience with generative AI techniques applied to LLM and multimodal learning (text, image, video).
• Knowledge of GPU/CPU architecture and related numerical software.

Compensation & Benefits

• Base salary ranges provided by location and level: 184,000 USD - 287,500 USD (Level 4); 224,000 USD - 356,500 USD (Level 5).
• Eligible for equity and benefits: see NVIDIA benefits information.

Other

• Applications for this job will be accepted at least until October 28, 2025.
• NVIDIA is an equal opportunity employer and committed to fostering a diverse work environment.

Relevant Open Source Projects Mentioned

• Megatron Core (Megatron-LM)
• NeMo Framework and NeMo-RL

#deeplearning