NVIDIA Launches DGX Cloud Lepton to Power Global AI Compute Marketplace

/ AI, Application Development, Cloud-Native, Infrastructure, Software Development
NVIDIA Launches DGX Cloud Lepton to Power Global AI Compute Marketplace

The News

At COMPUTEX 2025, NVIDIA introduced DGX Cloud Lepton™, a global compute marketplace designed to connect AI developers with tens of thousands of NVIDIA GPUs across a growing network of NVIDIA Cloud Partners (NCPs). Key partners include CoreWeave, Crusoe, Foxconn, SoftBank, Yotta Data Services, Lambda, GMI Cloud, and others. The platform supports Blackwell-based and other NVIDIA GPU architectures for AI workloads including generative, agentic, and physical AI applications. Learn more at nvidia.com.

Analysis

As AI use cases become more specialized, sovereign, and latency-sensitive, a federated compute infrastructure is necessary. DGX Cloud Lepton enables NVIDIA’s vision of a globally connected, regionally compliant, high-performance compute mesh — linking hyperscalers, sovereign clouds, and edge environments into one programmable AI infrastructure.

For developers, this means faster access to GPUs where and when they’re needed. For cloud providers, it offers a scalable path to join the NVIDIA ecosystem. For enterprises, it delivers consistent, secure, and performant AI compute anywhere innovation happens.

Solving the Global AI Compute Bottleneck

With the surge in large-scale AI development, securing reliable, on-demand access to performant GPUs has become one of the industry’s greatest constraints. DGX Cloud Lepton addresses this by:

  • Unifying access to regional and sovereign GPU resources
  • Offering flexible on-demand or long-term provisioning
  • Supporting cross-cloud, hybrid, and multi-region deployment requirements

Developers can tap into GPU capacity tailored to geographic or regulatory constraints, enabling scalable AI compute wherever workloads require it.

A New Paradigm for Distributed AI Factories

DGX Cloud Lepton introduces a federated compute model that supports:

  • Agentic AI training and inference workloads
  • Physical AI simulation and robotics
  • Multi-tenant development at cloud-native scale

This marketplace model effectively turns the global NVIDIA ecosystem into a planetary-scale AI factory, where compute is elastic, portable, and interoperable.

Full Stack Integration for Developer Productivity

DGX Cloud Lepton natively supports the NVIDIA AI software suite:

  • NVIDIA NIM and NeMo microservices for inference and LLM customization
  • NVIDIA Blueprints for AI system design patterns
  • NVIDIA Cloud Functions for event-driven AI application deployment

This integration ensures a seamless developer experience from prototyping to production across GPU clouds.

Enterprise-Class GPU Fleet Management for Cloud Providers

DGX Cloud Lepton provides NCPs with advanced operational tooling, including:

  • Real-time GPU health diagnostics
  • Automated root cause analysis for downtime reduction
  • Integration with NVIDIA Exemplar Clouds and DGX Benchmarking tools

This enhances reliability, performance predictability, and operational efficiency for cloud providers offering GPU-as-a-service.

Exemplar Clouds Set the Standard

To guide partners in delivering consistent enterprise-grade performance, NVIDIA also introduced Exemplar Clouds:

  • Reference designs built using DGX Cloud Benchmarking tools
  • Optimized security, usability, and workload tuning
  • First adopter: Yotta Data Services (Asia-Pacific region)

These clouds serve as blueprints for regional and global providers to meet the stringent demands of AI-native enterprises.

Author

  • Paul Nashawaty

    Paul Nashawaty, Practice Leader and Lead Principal Analyst, specializes in application modernization across build, release and operations. With a wealth of expertise in digital transformation initiatives spanning front-end and back-end systems, he also possesses comprehensive knowledge of the underlying infrastructure ecosystem crucial for supporting modernization endeavors. With over 25 years of experience, Paul has a proven track record in implementing effective go-to-market strategies, including the identification of new market channels, the growth and cultivation of partner ecosystems, and the successful execution of strategic plans resulting in positive business outcomes for his clients.

    View all posts