Technical Program Manager, Supercomputing
Core
Own compute allocation and optimize supercomputing resource usage for AI research teams.
Role type
Technical Program Manager (Supercomputing)
Builds
Compute allocation plans and utilization tools for AI research workloads
Domain
AI Research / High-Performance Computing
Deliverable
production ML models
Required skills
resource allocation, distributed systems, machine learning infrastructure, high-performance computing, tradeoff analysis, technical teaching, data investigation, stakeholder management
Preferred skills
compute allocation ownership at AI research lab, GPU clusters, distributed training, scheduling, workload placement, networking/storage/hardware capacity, capacity forecasting tools, mentoring researchers
Technologies
GPU clusters, distributed training, scheduling systems
Responsibilities
Own compute allocation and develop allocation plans with technical leads; Optimize resource usage by evaluating tradeoffs across capacity, hardware, and timing; Improve productive use of compute by investigating gaps between allocated resources and useful work; Teach teams to plan and use compute effectively; Connect research plans with infrastructure capacity; Learn from outcomes to improve allocation decisions
Seniority
Senior, hands-on IC