Dell Technologies is expanding its AI Data Platform with new enterprise context, GPU acceleration, and managed infrastructure capabilities to help organizations move AI workloads from pilot projects into production.
The updates include a Unified Semantic Layer, Enterprise Knowledge Graph, and knowledge agents designed to give AI systems governed business context, alongside NVIDIA-accelerated data processing, a new open-source storage benchmarking tool, and a fully managed PowerScale offering for Microsoft Azure.
For channel partners, the expansion also creates a broader opportunity in services around data strategy, governance, deployment, optimization, and ongoing AI infrastructure support as customers work through the operational challenges of putting AI into production.
- Dell adds enterprise context for agentic AI
- NVIDIA GPU acceleration speeds AI data processing
- Dell opens storage benchmarking for AI workloads
- Managed PowerScale expands in Microsoft Azure
- Dell to expand professional services for the AI Data Platform
- AI Data Platform creates services opportunity for partners
Dell adds enterprise context for agentic AI
Dell’s Unified Semantic Layer helps organizations define what their data means in a business context, while the Enterprise Knowledge Graph connects entities and relationships across enterprise systems.
The Unified Semantic Layer gives structured and unstructured information consistent business meaning through rules, definitions, and a searchable glossary.
For example, if one system refers to something as a client and another calls it an account, the layer can recognize that they represent the same thing. Organizations can also import existing ontologies and classification taxonomies, allowing them to reuse industry-standard or enterprise-specific taxonomy assets.
Enterprise Knowledge Agents build on that foundation to provide curated, permission-aware expertise and work with the AI models customers choose.
The Enterprise Knowledge Graph uses metadata, lineage, and query history to continually tune how data is connected. When an agent asks a question, the platform can pull in related information it is authorized to access, including tables, data products, multimodal data, and vector indexes, regardless of where they reside.
Knowledge Agents then put that context to work as trusted advisors focused on specific topics and grounded in defined portions of the Enterprise Knowledge Graph. Customers can determine what guidance an agent follows, what data it can access, what quality standards it must meet, and how much it is allowed to spend.
According to Varun Chhabra, Senior Vice President of Marketing for Dell’s ISG Solutions Group, these capabilities are intended to reduce the need for agents to repeatedly reconstruct context, lowering token consumption, and compute costs while improving the consistency and trustworthiness of AI-generated answers.
The capabilities also provide enterprises with greater flexibility to use smaller, locally hosted, or open-source models, with these models running within the customer’s own environment, including disconnected environments.
The Unified Semantic Layer, Enterprise Knowledge Graph, and Knowledge Agents remain within the customer’s data center, allowing the platform to build context from customer data and keep that context current as the business changes. The approach also gives customers flexibility across models, data, and storage providers rather than locking them into a single option.
Dell is also using NVIDIA Nemotron Retriever models within the platform for document parsing, embedding, and reranking. The models also provide reasoning and visual understanding capabilities to Knowledge Agents.
“What all of this drives is fewer steps and lower costs, because agents don’t have to keep rebuilding context in every request,” said Chhabra. “They actually have data that can flow into the agent, that reduces compute costs and the number of tokens that are generated. There is increased model choice.”
NVIDIA GPU acceleration speeds AI data processing
The announcement also incorporates NVIDIA accelerated computing, including NVIDIA cuDF, GPU-accelerated Apache Spark, and Apache Arrow, into the data processing environment.
The engine features up to 10x faster data processing, with jobs starting up to seven times faster over Apache Arrow.
The expansion of GPU acceleration across Dell’s data engines will speed up the preparation and processing of enterprise data for AI workloads.
Dell is also incorporating NVIDIA cuVS to accelerate vector indexing and search, extending GPU acceleration beyond data processing to help prepare and retrieve enterprise data for AI workloads.
Based on Dell’s internal benchmarking, Chhabra cited up to 20 times faster batch data processing compared with CPU-only processing and an average of four times faster data processing across workloads. Chhabra notes that the improvements depend on the workloads and configuration involved.
Dell’s testing was conducted using NVIDIA RTX PRO 4500 Blackwell Server Edition GPUs, with the Dell Data Processing Engine powered by NVIDIA cuDF.
Dell opens storage benchmarking for AI workloads
Further, Dell is introducing the Dell Storage Performance Benchmarking Tool, an open-source benchmarking tool for PowerScale and ObjectScale designed to help customers evaluate storage performance against their specific AI workloads.
The tool allows organizations to define and run realistic, repeatable storage benchmarks using their own infrastructure.
It supports testing scenarios such as checkpoint-style writes, high-concurrency reads, mixed read/write environments, and I/O queries. It measures throughput and latency in real time, verifies persistent data end-to-end, and records software versions and test provenance.
The tool can also test S3-compatible object storage across training, inference, and checkpointing workloads, helping organizations compare vendor solutions and determine whether their storage infrastructure can keep GPUs supplied with data.
The release of the tool helps customers make more informed storage-sizing decisions by using performance results from their own environments rather than relying exclusively on standardized vendor benchmarks.
Managed PowerScale expands in Microsoft Azure
Dell has also announced the expansion of its PowerScale for Microsoft Azure, which is offered as a fully managed service, allowing customers to use PowerScale in Azure without handling deployment, monitoring, maintenance, or upgrades themselves.
PowerScale’s multitenancy and security features include up to 500 tenants per cluster, mTLS over NFS, and a more granular RBAC.
The offering was previously available as a customer-managed capability, but now new Dell-managed options shift those operational responsibilities to Dell, while Microsoft provides the underlying cloud infrastructure. It doesn’t require customers to purchase or lease an appliance.
Dell to expand professional services for the AI Data Platform
Also announced: Dell is expanding its professional services portfolio to help enterprises plan, deploy, and optimize their AI Data Platform environments.
The expanded services cover data strategy, production-ready platform implementation, data engine selection, and ongoing platform optimization.
The services are designed to help customers determine where their data resides, identify appropriate data sources, align infrastructure with AI use cases, and transition AI projects from pilots to productions.
Dell’s implementation services also extend across its data and storage engines, activating the platform’s analytics, processing, search, and orchestration capabilities while keeping the platform tuned to the demands of AI workloads.
“These new capabilities help organizations move from platform deployment to production results quickly and reliably,” said Chhabra. “The professional services here support the full AI Data Platform lifecycle. They help customers establish a data strategy before you even start talking about infrastructure.”
Dell’s managed services simplify ongoing operations, as well as deployment and support services.
AI Data Platform creates services opportunity for partners
Chhabra notes that the AI Data Platform announcements can create opportunities for channel partners to expand their services portfolio beyond traditional resale.
Partners can provide value-added services that help customers establish data strategies, evaluate data sources, address governance and legal considerations, and connect fragmented data environments.
“The opportunities for the channel ecosystem beyond reselling is really around providing those value-added capabilities, and I think that what’s happening with AI provides a massive inflection point for channel partners to really look at their services offerings and be able to really drive value across all parts of the daily lifecycle through consulting through professional services through support and deployment,” said Chhabra.
It’s an opportunity for channel partners to reassess their business models and develop capabilities in consulting, professional services, deployment, and support. Chhabra stressed that helping customers manage data and AI adoption is an ongoing process rather than a one-time implementation.