AI News

Vertiv Opens ATLAS Lab to Test AI Infrastructure with Simulation Before Chips Arrive

Unite.AIUpdated Nov 16 22:38

AI summary

Vertiv announced on November 16, 2026 that the ATLAS Lab will test the power, cooling, control and IT systems of an AI factory as a single integrated installation, and will use simulated NVIDIA compute to expose integration issues before production systems come to market. The lab corresponds to one AI factory pod, with testing extending from grid connection to the rack and tracking the entire process from chip-level liquid cooling to heat rejection and reuse. ATLAS also combines the Omniverse libraries, DSX Sim and SimReady assets of the NVIDIA DSX AI Factory Platform for digital twin validation, with the goal of turning pod-level validation results into campus-level reference designs.

Body

Vertiv announced on November 16, 2026 that its ATLAS Lab will test AI factory power, cooling, controls and IT systems as one installation, using emulated NVIDIA computing before production systems reach the market. The goal is to expose integration problems during engineering rather than at a customer’s site.

ATLAS stands for Advanced Test Lab for Architecture and Systems. Vertiv and NVIDIA engineering leaders held a ribbon-cutting and technical working sessions on November 6, 2026, moving from architecture requirements and digital simulation to assessments of how power, liquid cooling, controls and emulated AI workloads behave together.

The lab addresses a mismatch in validation: infrastructure equipment is commonly assessed product by product, while an AI factory functions as a connected system. Vertiv identifies five pressures: density, speed, scale, complexity and dynamic load profiles. Those pressures can reinforce one another, so solving one independently may shift the constraint to another part of the installation.

The physical installation represents one AI factory pod, with electrical, thermal, control and IT equipment in the same environment. Vertiv says its workload emulation reproduces the power and heat behavior of current and future NVIDIA accelerated-computing architectures, letting engineers test infrastructure responses before the corresponding production compute systems are commercially available.

Engineers first use simulation and form-and-fit testing to assess design risk. They then emulate future compute loads and stress the power and cooling systems together, with coordinated controls responding to rapid, synchronized AI workload swings. Test coverage begins at the grid connection, continues to the rack, and follows heat from chip-level liquid cooling through rejection and reuse.

Vertiv says a design validated at pod scale can become a reference design for a campus, avoiding the need to re-engineer infrastructure for each deployment. That makes the pod more than a hardware testbed: it is intended to establish a repeatable technical baseline for customer projects.

ATLAS combines those physical tests with simulation and digital-twin workflows in the NVIDIA DSX AI Factory Platform, using Omniverse libraries, DSX Sim and SimReady assets. Models incorporate Vertiv power and cooling assets so infrastructure designs can be evaluated before physical deployment, with requirements informed by NVIDIA’s AI factory reference designs.

NVIDIA describes DSX as unifying design, simulation, operations and ecosystem technologies across grid, facilities, chips, infrastructure software and partner systems. DSX Sim specifically models facility, infrastructure and IT systems together before construction. Within ATLAS, those digital tools are paired with physical integration work covering power architecture, liquid cooling and controls.

Integration Across Suppliers

Vertiv is positioning the lab as a common environment for chip manufacturers, server and rack-system makers, power and cooling providers, utilities and operators. Their technologies meet at shared interfaces, where integration problems can affect design, deployment and operation. A common test setting gives those parties a shared technical basis for decisions across chips, racks, facilities and the grid.

Vertiv chief executive Giordano Albertazzi said ATLAS allows the company to “prove the complete architecture before a customer commits capital to build it.” NVIDIA vice president of AI infrastructure Vladimir Troy said earlier system-level validation can help resolve infrastructure challenges before compute systems arrive.

The companies’ stated objective is to shorten time to first token and turn available megawatts into productive compute sooner. The business significance therefore rests on repeatability: whether a validated pod can supply a campus-scale design that reduces site-by-site integration work and provides a tested starting point for deployment.

Source

Unite.AI