GPU server racks representing AI factory power management.

End-to-End AI Factory Efficiency

Turn Every Megawatt Into More Intelligence.

Plan, deploy, and operate AI factories that maximize productive capacity within fixed power, water, and grid constraints.

Explore NVIDIA DSX

Overview

Optimize the AI Factory

The NVIDIA DSX™ AI factory platform combines accelerated computing with workload-aware power orchestration, high-voltage distribution, direct liquid cooling, digital twins, and fleet operations to maximize tokens, capacity, and economic value per megawatt.

Video

More AI From the Power You Have

See how NVIDIA DSX connects power, cooling, compute, and workloads to help AI factories deliver more capacity with greater energy efficiency.

YouTube Video

Benefits

How DSX Improves AI Output per Megawatt

From power envelopes to rack-scale optimization, system design directly drives efficiency.

Maximizes Performance per Watt

NVIDIA DSX MaxLPS™ dynamically manages power across GPUs, racks, and workloads to maximize useful compute.

Enables Higher Rack Density

800 VDC architecture reduces power conversion stages and distribution losses to support next-generation accelerated computing and more compute density within racks.

Lowers Cooling Overhead

Direct-to-chip liquid cooling enables high-density deployments while significantly lowering cooling overhead.

Responds to Grid Conditions

DSX Flex coordinates workloads with site and grid power conditions to preserve performance and grid stability.

Technologies

DSX Technologies for Energy Efficiency

Design, power, cooling, and operations technologies work together to produce more AI within fixed power budgets.

NVIDIA DSX MaxLPS

Optimize AI factory power efficiency with MaxLPS to monitor and allocate power based on utilization across GPU, rack, and cluster levels to maximize AI throughput within strict power budgets.

Explore MaxLPS Power Optimization

800 VDC Architecture

Support future generations of accelerated computing and rack power delivery while reducing conversion stages and distribution losses.

Get to Know 800 VDC Architecture

NVIDIA DSX Flex

Dynamically coordinate AI workloads with grid conditions to optimize power use and support grid reliability.

Discover DSX Flex Power Orchestration

NVIDIA DSX Sim™

Model facility, infrastructure, and IT systems together before construction to validate layouts, operations, and design assumptions.

Simulate an AI Factory

NVIDIA DSX Exchange™

Open source, real-time event bus connecting data center facility telemetry with AI workload management for dynamic power orchestration.

View DSX Exchange Docs

Resources

The Latest From the NVIDIA DSX Ecosystem

See All Tech Blogs See All Topic News

DSX Overview Docs

Explore the features and capabilities of the DSX platform.

Learn More

DSX MaxLPS Docs

Learn how to design and operate AI factories for more throughput per megawatt.

Learn More

DSX Flex Docs

Learn how to enable renewable and hybrid power orchestration across utility, on-site renewables, and storage.

Learn More

DSX Sim Docs

Explore the suite of simulation technologies that together enable partners to design, validate, and operate AI factories before and after physical deployment.

Learn More

DSX Exchange Docs

Learn more about how DSX Exchange is an open source event bus for AI factory operations.

Learn More

DSX OS Docs

Learn more about open source, modular software for building, operating, and scaling multi-tenant AI factory infrastructure.

Learn More
See All Docs

Next Steps

Start Building With NVIDIA DSX

Design and optimize AI factories with NVIDIA reference architectures, simulation tools, and ecosystem software.

Browse DSX Documentation
Products
Technologies
Resources
Company Info