As enterprise AI adoption accelerates and the volume of data multiplies, organizations need a platform that can securely transform distributed information into actionable intelligence. Therefore, Dell Technologies has announced new and significant advances in its AI Data Platform. Designed to help businesses quickly convert their distributed and fragmented data into reliable and agile AI results, this platform is a key component of the Dell AI Factory initiative.
“AI is transforming industries, and its success depends on unlocking the full potential of enterprise data. The Dell AI Data Platform is specifically designed to simplify data complexity, unify pipelines and deliver AI-ready data at scale,” explains Arthur Lewis, president of the Infrastructure Solutions group at Dell Technologies.
For his part, Justin Boitano, vice president of enterprise AI products at NVIDIA, has added important perspective on the strategy by stating, “AI offers companies a way to transform fragmented data into a strategic and scalable asset. Accelerated by NVIDIA AI, Dell AI Data Platform offers a new generation of intelligent storage that is designed to understand the meaning behind the data it contains.”
This modular design helps eliminate common bottlenecks by decoupling data storage from processing, offering the flexibility to handle demanding AI workloads, including training, fine-tuning, inference, or recovery-augmented generation (RAG).
The Foundation, Storage and Integrated Cyber Resilience
The platform, integrated with the NVIDIA AI Data Platform reference design, is built on four fundamental pillars, including storage engines for the intelligent placement and seamless movement of data, data engines that transform data into actionable information, integrated cyber resilience, and data management services.
“AI is transforming industries, and its success depends on unleashing the full potential of enterprise data”
The storage engines, Dell PowerScale and Dell ObjectScale, are essential to providing the performance, security, and multi-protocol access required to handle massive volumes of AI data. Dell PowerScale delivers the signature simplicity of NAS storage along with parallel performance vital for AI workloads like RAG pipelines.
With the integration of NVIDIA GB200 and GB300 NVL72 GPUs, PowerScale ensures reliable performance and simplified management at scale. In terms of efficiency, the NVIDIA Cloud Partner (NCP) certified PowerScale F710 provides performance for more than 16,000 GPUs, using up to 5% less rack space and up to 72% less power consumption compared to the competition.
Considered the industry’s highest-performing object platform, ObjectScale offers native S3 object storage that is extremely fast and scalable for massive AI workloads. A new software-defined option on Dell PowerEdge servers enables up to eight times the performance of object storage all-flash of the previous generation. Additionally, support for S3 over RDMA, which will enter technical preview, promises significant improvements, such as up to 230% higher throughput and 80% lower latency.
Data Engines, Analytics Drivers and Vector Search
Dell is expanding the capabilities of its data engines, the specialized tools that organize, query and activate information for AI. These engines are the result of collaboration with leaders such as NVIDIA, Elastic and Starburst.
The new Data Search Engine, developed with Elastic, significantly accelerates decision-making by allowing customers to interact with data naturally, asking direct questions. Designed for RAG and semantic search, it integrates with MetadataIQ software to search billions of files stored in PowerScale and ObjectScale using granular metadata.
For its part, the Data Analytics Engine, created in collaboration with Starburst, allows seamless data queries across databases, clouds, spreadsheets and lakehouses. Its new Agentic Layer uses language models (LLM) to automate documentation and draw insights, integrating AI into SQL workflows in seconds. This tool also unifies access to vector stores, enabling search and RAG tasks in diverse environments.
A crucial advance is the integration of the Dell AI Data Platform with NVIDIA cuVS, which represents a leap in vector search performance. This integration offers a “turnkey” implementation for enterprise environments, with hybrid search (by keywords plus vector) accelerated by GPU, ensuring faster and more efficient information with full control on-premise.
