Dell Data Orchestration Engine joins AI data pipeline fray
Dell jumps into the ring with NetApp and Vast Data, unveiling a new AI data orchestration product built on its Dataloop acquisition as Nvidia STX shakes up the storage industry.
Enterprise IT buyers grappling with reining in AI data have a new low-code / no-code pipeline builder and data orchestration tool to consider from Dell, as well as a plethora of enterprise storage systems supporting a new Nvidia architecture.
This week at Nvidia's GTC conference, Dell made its entry into the AI data orchestration market with a new Data Orchestration Engine, based on IP it acquired with Dataloop in December. That unveiling was part of a broader Dell data management and storage product refresh that also included the general availability of Dell's long-awaited Lightning FS high-performance parallel file system and a new data storage system called Exascale that combines file, object and block support.
"Data Orchestration Engine is a low-code / no-code engine that automatically discovers, prepares and governs structured and unstructured data, as well as multimodal data, into AI-ready data sets at scale," said Varun Chhabra, senior vice president of Dell's Infrastructure Solutions Group, during a press briefing this week.
"Six months ago … our capabilities were more based on a specific engine, such as vector search analytics sitting on top of our storage platforms," Chhabra added. "But what was still missing was the ability to take the data across the entire estate that the [Dell] AI Data Platform provides and … automate various tasks related to that data wherever customers are."
Nvidia STX charges up AI data storage
The Dell Data Orchestration Engine fits into its broader Dell AI Data Platform, which Dell also refreshed this week with support for Nvidia's latest storage processors and accelerators, dubbed the Nvidia STX architecture. Nvidia STX includes the Nvidia BlueField-4 data processor that combines a Vera CPU with its ConnectX-9 SuperNIC. STX also integrates Nvidia Spectrum-XEthernet networking, DOCA software framework and AI Enterprise software. Storage vendors' support for Nvidia tools also commonly includes support for the Nvidia libraries that accelerate data manipulation on GPUs -- cuDF -- and vector search -- cuVS.
Data storage vendors flocked en masse this week to support STX, including Cloudian, DDN, Everpure, Hitachi Vantara, HPE, IBM, MinIO, NetApp, Nutanix, Vast Data and WEKA.
"One hundred percent of the world's storage industry is joining us on this system," said Nvidia CEO Jensen Huang, during a GTC keynote presentation this week. "And the reason for that is because they see exactly the same thing: the storage system is going to get pounded.
The cuDF and cuVS integrations quietly showing up inside Snowflake, Starburst, watsonx -- those matter more to an enterprise IT leader's next 12 months than anything Jensen [Huang] showed on the big stage.
Mike LeoneAnalyst, Omdia
"It's going to get pounded because we used to have humans using the storage systems … using SQL. Now we're going to have AIs using these storage systems, and it's going to store cuDF-accelerated storage, cuVS-accelerated storage, as well as, very importantly, KV caching."
STX slots into the behemoth Vera Rubin NVL72 rack-scale system, which is still mostly meant for hyperscalers and neoclouds, though practically every major enterprise vendor also pledged support for it in AI Factory products this week.
But STX and ubiquitous cuDF and cuVS integration stand to have a more significant long-term effect on enterprise AI, said Mike Leone, an analyst at Omdia, a division of Informa TechTarget.
"The cuDF and cuVS integrations quietly showing up inside Snowflake, Starburst, watsonx -- those matter more to an enterprise IT leader's next 12 months than anything Jensen [Huang] showed on the big stage," he said. "Practically speaking, it means GPU acceleration will just show up inside the data platforms enterprises already run. Queries will run faster, vector search will get cheaper, and the ROI will show up for those existing workloads. In other words, this isn't a benefit just for new AI projects."
AI blurs data storage, management, app platform lines
Data management for AI applications has emerged as a top challenge for enterprise AI, along with security and governance, as early-stage pilot projects struggle to reach production at many companies. Data quality and security were cited as the top concern for building and operating bespoke agentic AI platforms within their organization, among 894 respondents to IDC's 2025 Future Enterprise Resiliency and Spending Survey.
Respondents could choose up to two responses from a menu of concerns, and "data quality and security" was chosen by 49.5%. The next most chosen concerns, "integration with existing systems" and "high costs of implementation" were picked by 34.3% and 31.6%, respectively.
"I'm seeing a convergence of data platforms and application platforms," said Matthew Flug, an analyst at IDC. "These platforms, and the people who manage them, are now responsible for making those secure connections to data, and they're going to need to be on [enterprise AI] platform teams because they're the ones who know the data the best."
While application platform vendors such as Broadcom have extended into data management, data management vendors such as Snowflake have begun reaching into the application layer with moves like the acquisition of Observe, Flug said. And further down the stack, enterprise data storage vendors such as NetApp, Vast Data and Everpure have begun moving upward into data intelligence and orchestration.
Dell's approach with its Data Orchestration Engine stands in contrast to NetApp's AI Data Engine, which requires NetApp's OnTap OS, according to Rob Strechay, an analyst at TheCube Research and Smuget Consulting.
"A major difference is that the Dell Data Orchestration Engine will work beyond its own storage," Strechay said. "It works at the technical metadata and up a bit into the business metadata layer, although this is thin right now and mainly with partners."
The new engine also represents a significant competitive move against Vast, Strechay said, which has pivoted from software-defined data storage into AI-native infrastructure over the last year.
"Dell would say their tech goes beyond what Vast is doing in its own walled garden," he said. "This is a game of leapfrog at the moment, with both Dell and Vast quickly adding features to move from the storage and storage services layers, into the data services, GRC [governance, risk and compliance] and metadata layers. The big question is: can either company be successful selling to personas it hasn't traditionally sold to?"
Beth Pariseau, a senior news writer for Informa TechTarget, is an award-winning veteran of IT journalism. Have a tip? Email her.
