Maximizing AI Factory Performance per Watt with NVIDIA DSX MaxLPS | NVIDIA Technical Blog
Aug 24, 2026Tanya Lenz

Summary
This post details NVIDIA DSX MaxLPS, a suite of technologies designed to maximize AI factory throughput within a fixed power budget. It introduces Dynamic Power Software (DPS) for real-time monitoring and reallocation of unused power headroom across managed groups, and workload profile power solutions via the Application Performance and Power Manager to align GPU configuration with specific workload types. It also highlights the thermal efficiency gains from Vera Rubin NVL72 racks operating at 45° C, enabling increased free cooling and improved PUE. The post quantifies potential gains, projecting up to 40% more GPU capacity within a 100 MW AI factory through these optimizations.
More from the AI Factory Power Management and Optimization trail
- How NVIDIA DSX MaxLPS Maximizes AI Factory Throughput and Efficiency | NVIDIA Technical Blog
- How NVIDIA Groq 3 LPX Deterministic Execution Drives Power-Efficient High-Interactivity Inference on NVIDIA Vera Rubin | NVIDIA Technical Blog
- Maximize AI Factory Energy Efficiency Through Full-Stack Inference and Training Optimizations | NVIDIA Technical Blog
- Delivering Lifecycle Control for AI Infrastructure at Scale with NVIDIA DGX Spark Enterprise Manageability | NVIDIA Technical Blog
- NVIDIA DSX OS Delivers Open, Modular Software for Operating AI Factories at Scale | NVIDIA Technical Blog