Research Engineer - Multimodal Training Infrastruc
Core
Research and develop large-scale infrastructure for training foundation models, multimodal LLMs, and image/video generation models.
Role type
Senior IC research engineer (multimodal training infrastructure)
Builds
Scalable, production-ready distributed training systems for foundation models
Domain
AI/ML infrastructure, large-scale distributed systems
Deliverable
production ML models | infrastructure
Required skills
large-scale distributed training, data/model/pipeline/expert parallelism, performance optimization on GPU clusters, algorithm–system co-design, fault tolerance, checkpointing, cross-layer optimization
Preferred skills
exascale system analysis, data-driven optimization methods
Technologies
GPU clusters, distributed training frameworks
Responsibilities
Design and optimize distributed training strategies; Improve computation, communication, throughput, networking, scheduling, and GPU memory management; Research reliability and resilience techniques; Analyze performance bottlenecks in exascale training systems; Translate research ideas into scalable, production-ready infrastructure solutions
