NVIDIA DSX Gives Infrastructure Builders the Playbook for AI Factories

Information Abstract:

• Engineered from the bottom up for AI factories, the NVIDIA DSX platform defines how next-generation infrastructure is designed, constructed and operated — driving lowest token value and accelerating time to first manufacturing throughout NVIDIA chips, methods, software program, amenities and associate applied sciences.
• New DSX MaxLPS software program allows AI infrastructure and factories to ship the bottom token value by maximizing token efficiency per megawatt.
• Open supply, modular DSX OS software program brings collectively lifecycle administration, runtime consistency and well being automation, resiliency, multi-tenant AI manufacturing unit operations and platform providers.
• Trade-leading producers are constructing NVIDIA DSX-ready methods supporting the buildout of AI factories with excessive codesign.
• Rising DSX partnerships throughout each layer of the stack speed up the design, deployment and operation of AI factories worldwide.

TAIPEI, Taiwan, Could 31, 2026 (GLOBE NEWSWIRE) — NVIDIA GTC Taipei — NVIDIA right this moment introduced the NVIDIA DSX™ platform, which supplies infrastructure builders a whole playbook to create AI factories.

NVIDIA DSX brings collectively open supply, modular software program libraries, software programming interfaces, reference designs, NVIDIA accelerated computing platforms and associate applied sciences into a typical, codesigned platform for AI manufacturing unit design, deployment and operations.

NVIDIA is the one firm that builds the full AI factory. By aligning each layer of the stack throughout compute, software program, amenities and associate applied sciences, DSX supplies infrastructure builders with a confirmed framework to design, deploy and function AI factories at scale.

The built-in platform accelerates deployment, improves operational reliability and resiliency at scale and allows a broad ecosystem of options designed to show each megawatt into extra intelligence on the lowest token value.

“We’re not simply transport chips — we’re giving each infrastructure builder a whole playbook to construct AI factories,” stated Jensen Huang, founder and CEO of NVIDIA. “With the DSX platform, you may simulate all the manufacturing unit earlier than you spend a greenback, validate efficiency earlier than a single rack is put in and function with the form of reliability that manufacturing AI calls for.”

DSX Platform Parts
DSX now spans the total stack, from silicon and methods to infrastructure software program, amenities and associate applied sciences. The newest additions to the platform embody new open supply software program:

• DSX MaxLPS™: A collection of applied sciences to maximise token efficiency per megawatt inside a hard and fast energy price range, enabling lowest token value for AI factories. Combining 45-degrees-Celsius liquid cooling with in-rack applied sciences that optimize efficiency per watt, DSX MaxLPS lets operators run as much as 40% extra GPUs at their most energy-efficient working level with minimal affect on workload efficiency.
• DSX OS™: Open supply, modular software program purpose-built for AI manufacturing unit operations, offering lifecycle administration, intelligence scheduling, runtime consistency, well being automation, resiliency, multi-tenant operations and platform providers.

DSX MaxLPS and DSX OS be a part of an present set of options below the DSX platform:

• DSX Reference Design: Technology-specific, validated AI manufacturing unit architectures overlaying compute, networking, storage, {hardware} cluster design and amenities infrastructure — together with energy, cooling and controls, in addition to civil, structural and architectural design.
• DSX Sim™: Excessive-fidelity simulation layer for the AI manufacturing unit lifecycle, serving to NVIDIA, companions and prospects to mannequin, validate and optimize infrastructure choices from planning and design via deployment and operations.
• DSX Flex: Connects AI factories to power-grid providers, enabling dynamic workload adaptation to grid indicators corresponding to load shedding, demand response and pricing occasions, and orchestrating renewable and hybrid energy throughout utility, onsite renewables and storage.
• DSX Change™: Allows scalable, safe integration of compute, community, vitality, energy and cooling plant indicators between IT, operational expertise and operations brokers.

Rising DSX Ecosystem
NVIDIA is partnering with industry-leading Taiwan system producers to expand the DSX ecosystem, supporting the buildout of AI factories with excessive codesign at their core.

NVIDIA cloud companions CoreWeave, Crusoe, Firmus, IREN, Lambda, Nebius, Nscale and Yotta Knowledge Companies are deploying core elements of the DSX platform stack — DSX Sim, DSX MaxLPS and DSX OS — to scale back danger, enhance GPU utilization and convey AI cloud capability on-line quicker.

Dell Applied sciences, HPE, Lenovo and Supermicro along with ASUS, Foxconn, GIGABYTE, Pegatron, Quanta Cloud Expertise (QCT), Wistron and Wiwynn are constructing NVIDIA DSX-ready methods and contributing simulation-ready property that allow prospects to deploy full, full-stack AI manufacturing unit options at world scale.

Throughout the ecosystem, model-based methods engineering serves because the bridge between rack design to facility deployment, for an AI infrastructure optimized for token efficiency per megawatt. Quanta Cloud Expertise (QCT) and Pegatron are working with Dassault Systèmes to create a stay AI manufacturing unit digital twin configurator to automate rack-to-facility design with elevated high quality and diminished workload. The adoption of DSX Sim by system producers expands the NVIDIA Omniverse DSX Blueprint ecosystem, deepening integration with software program companions Cadence, PTC and Siemens.

DSX Flex is powering a industrial, multi-megawatt pilot with Emerald AI and Silicon Valley Energy to reveal grid-responsive AI factories that may dynamically regulate energy consumption in response to utility indicators whereas defending AI workload efficiency, serving to safeguard grid reliability and affordability for patrons whereas unlocking further energy capability to assist AI progress.

Companions are adopting varied DSX OS software program elements for lifecycle administration, multi-tenancy, safety, well being automation, resilience and platform providers. Ecosystem companions adopting DSX OS elements embody Aible, BeyondAI, Bhashini, DCAI, Mirantis, OpenNebula Programs, Rafay, Crimson Hat, Sarvam, Simplismart, Spectro Cloud, Supermicro, vCluster and Vultr.

Watch Huang’s keynote and study extra at NVIDIA GTC Taipei.

About NVIDIA
NVIDIA (NASDAQ: NVDA) is the world chief in AI and accelerated computing.

For additional data, contact:
Kristin Uchiyama
Company Communications
NVIDIA Company
press@nvidia.com

Sure statements on this press launch together with, however not restricted to, statements as to: expectations with respect to progress, efficiency, availability, and advantages of NVIDIA’s merchandise, providers and applied sciences, and associated traits and drivers; expectations with respect to NVIDIA’s third celebration preparations, together with with its collaborators and companions; expectations with respect to expertise developments, and associated traits and drivers; projected market progress and traits; expectations with respect to AI and associated industries; and different statements that aren’t historic details are forward-looking statements throughout the which means of Part 27A of the Securities Act of 1933, as amended, and Part 21E of the Securities Change Act of 1934, as amended, that are topic to the “secure harbor” created by these sections based mostly on administration’s beliefs and assumptions and on data presently out there to administration and are topic to dangers and uncertainties that would trigger outcomes to be materially completely different than expectations. Essential components that would trigger precise outcomes to vary materially embody: world financial and political circumstances; NVIDIA’s reliance on third events to fabricate, assemble, package deal and take a look at NVIDIA’s merchandise; the affect of technological improvement and competitors; improvement of latest merchandise and applied sciences or enhancements to NVIDIA’s present merchandise and applied sciences; market acceptance of NVIDIA’s merchandise or NVIDIA’s companions’ merchandise; design, manufacturing or software program defects; adjustments in client preferences or calls for; adjustments in {industry} requirements and interfaces; surprising lack of efficiency of NVIDIA’s merchandise or applied sciences when built-in into methods; NVIDIA’s capability to comprehend the potential advantages of enterprise investments or acquisitions; and adjustments in relevant legal guidelines and laws, in addition to different components detailed sometimes in the latest experiences NVIDIA information with the Securities and Change Fee, or SEC, together with, however not restricted to, its Annual Report on Kind 10-Okay and Quarterly Stories on Kind 10-Q. Copies of experiences filed with the SEC are posted on the corporate’s web site and can be found from NVIDIA with out cost. These forward-looking statements will not be ensures of future efficiency and converse solely as of the date hereof, and, besides as required by legislation, NVIDIA disclaims any obligation to replace these forward-looking statements to replicate future occasions or circumstances.

Most of the merchandise and options described herein stay in varied phases and will likely be provided on a when-and-if-available foundation. The statements above will not be supposed to be, and shouldn’t be interpreted as a dedication, promise, or authorized obligation, and the event, launch, and timing of any options or functionalities described for our merchandise is topic to alter and stays on the sole discretion of NVIDIA. NVIDIA may have no legal responsibility for failure to ship or delay within the supply of any of the merchandise, options or features set forth herein.

© 2026 NVIDIA Company. All rights reserved. NVIDIA, the NVIDIA emblem, DSX Change, DSX OS, DSX Sim, DSX MaxLPS and NVIDIA DSX are logos and/or registered logos of NVIDIA Company within the U.S. and/or different nations. Different firm and product names could also be logos of the respective firms with which they’re related.

A photograph accompanying this announcement is out there at https://www.globenewswire.com/NewsRoom/AttachmentNg/ab7c906c-4467-47b2-942a-b09fcd152f65