Dell Technologies Inc. unveiled a sweeping set of updates to its artificial intelligence infrastructure portfolio today, adding new data platform capabilities, AI-optimized storage, servers, networking and workstation systems designed to help enterprises move AI projects from experimentation into large-scale production.

The announcements, made at Nvidia Corp.’s GTC conference, mark the two-year anniversary of the Dell AI Factory with Nvidia, a joint initiative aimed at providing full-stack infrastructure for enterprise AI training and inference. Dell executives said the platform has already gained significant traction, with more than 4,000 customers.

Today’s announcements span more than 30 product and technology updates across data platforms, storage, servers, workstations, networking and services. Dell said the goal is to simplify the complex infrastructure stack required to operationalize AI while addressing the biggest deployment obstacles.

Unclogging the data bottleneck

Feedback from those 4,000 customers has revealed that data, not infrastructure, is the biggest choke point, said Varun Chhabra, senior vice president of product marketing for Dell’s Infrastructure Solutions Group. “Data availability and quality are the most common challenges in AI implementations among both high-maturity and low-maturity organizations,” he said.

That’s followed by the need to deploy models across increasingly varied infrastructure configurations.

“Third-party analysis shows that by 2028, 75% of enterprise AI workloads will be deployed on hybrid infrastructure,” Chhabra said. “That is being done to optimize performance, cost and compliance.”

A third challenge is identifying the business use cases that deliver measurable returns. “Without an expert or a trusted partner at their side, organizations can often struggle to choose between different use cases,” Chhabra said. “That ends up leading to unclear ROI in many AI projects.”

Dell is targeting all three areas with enhancements to its AI Data Platform with Nvidia, which combines Dell storage technologies with Nvidia-accelerated computing and software.

The updated platform integrates Dell storage systems such as PowerScale and ObjectScale with new data engines and a data orchestration layer designed to automate the preparation of structured and unstructured datasets.

“The Dell AI Data Platform with Nvidia provides a unified, vertically integrated data foundation that brings analytics and AI together in a single stack,” Chhabra said.

A new component is the Dell Data Orchestration Engine, a low-code system designed to automate data discovery, preparation and governance. It automatically discovers, prepares, governs and structures structured, unstructured and multimodal data into AI-ready data sets at scale, Chhabra said.

The platform also introduces a conversational AI assistant integrated into Dell’s analytics engine, allowing business users to query governed datasets using natural language rather than SQL.

Dell said performance improvements from the platform include faster vector indexing, accelerated data processing and reduced AI inference latency.

Storage for large AI workloads

Dell also announced new storage technologies aimed at supporting increasingly demanding AI workloads.

Among the most notable is the Dell Lightning File System, the commercial release of a technology previously known as Project Lightning, a high-performance parallel file system designed to accelerate AI model training and inferencing.

“The Lightning File System delivers tier-0 ultra-high-performance storage purpose-built for the most demanding AI workloads,” Chhabra said. “We believe it is the world’s fastest parallel file system.”

The system is designed to maximize graphic processing unit utilization by preventing storage bottlenecks that can stall training and inference pipelines.

Dell also introduced Dell Exascale Storage, a software-defined architecture that allows organizations to deploy file, object and parallel file storage on the same infrastructure.

The platform is “not a new storage product, but rather a storage delivery vehicle that customers can use to adapt different storage personalities over time without having to replatform,” Chhabra said. The platform targets extremely large deployments, including AI clusters larger than 10 petabytes.

Dell also unveiled multiple server updates designed for AI workloads in data centers. The PowerEdge XC9812 is a high-end liquid-cooled server built on Nvidia’s Vera Rubin NVL72 platform. Three other new liquid-cooled models — the XC9880L, XC9885L and XC9882L — are designed for dense GPU deployments. The rack designs allow up to 144 Rubin GPUs to be packed into every rack.

For more mainstream enterprise AI workloads, Dell announced new PowerEdge R770, R7715 and R7725 servers using Nvidia RTX Pro 4500 Blackwell GPUs. The company also unveiled PowerEdge systems supporting Nvidia’s Vera CPUs, enabling large-scale deployments of Arm Holdings Plc-based CPUs for AI, analytics and high-performance workloads.

Networking and hybrid quantum

Dell is expanding its networking portfolio with new PowerSwitch SN6000 series switches, which integrate Nvidia Spectrum-6 Ethernet technology and deliver 1.6 terabit-per-second bandwidth with optional liquid-cooling.

Chhabra said the design reduces power consumption and improves reliability for AI networks. “The secret sauce is the co-packaged optics integrating silicon photonics directly with the switch [application-specific integrated circuit],” Chhabra said. The company also announced support for Nvidia’s Quantum-X800 InfiniBand switches for high-bandwidth AI fabrics.

Another notable development is Dell’s work on hybrid quantum-classical computing, combining Nvidia GPUs with quantum processing units. Dell is integrating Nvidia’s NVQLink and Cuda-Q hybrid quantum-classical computing platform into PowerEdge servers to support advanced simulations and emerging quantum use cases.

“Hybrid quantum-classical computing will be a game changer for industries such as pharmaceutical research,” Chhabra said.

Packaged AI software

Finally, Dell is rolling out new packaged AI software components, including knowledge assistants and an agentic AI platform built with partners such as Cohere Inc. and DataRobot Inc.

“The goal is to provide customers with a turnkey system already validated for the latest Nvidia technologies, and then give them an easy path to scale compute, networking and storage independently as their deployments grow,” Chhabra said.

With more than 30 new announcements and more than 240 updates introduced since the launch of the AI Factory initiative two years ago, Dell executives said, the company’s focus is on enabling organizations to put AI into operation at large scale.

“We’ve been on a tear in terms of innovation speed,” Chhabra said.

Image: Dell/X