Dell and NVIDIA framed enterprise Artificial Intelligence as moving beyond pilots and into large-scale agentic and inference deployments. Michael Dell said a major investment cycle is already underway, while Jensen Huang described demand as accelerating rapidly as useful Artificial Intelligence reaches production use. The companies positioned the Dell Artificial Intelligence Factory with NVIDIA as the platform for running frontier models and autonomous agents securely behind the enterprise perimeter, with a focus on productivity gains, faster iteration and heavier computation requirements.
The hardware announcements centered on new infrastructure for the agentic era. The Dell PowerEdge XE9812, built on NVIDIA Vera Rubin NVL72, delivers up to 10x lower cost-per-token than NVIDIA Blackwell for massive-scale agentic Artificial Intelligence inferencing. Dell also introduced PowerEdge XE9880L, XE9885L and XE9882L servers built on NVIDIA HGX Rubin NVL8, supporting up to 144 GPUs per rack with 100% direct liquid-cooled compute nodes and up to 5.5x the performance of HGX B200. On the CPU side, Dell PowerEdge M9822 and R9822 servers bring NVIDIA Vera CPUs to enterprise Artificial Intelligence factories, where Vera uses 1.2 TB/s memory bandwidth and completes agentic workloads 50% faster than x86 processors. Starburst, part of the Dell Artificial Intelligence Data Platform with NVIDIA, delivers 3x faster query throughput on NVIDIA Vera CPU for large-scale SQL analytics.
Dell also emphasized that enterprise adoption is increasingly happening outside public cloud environments. Its survey data cited onstage found that 67% of Artificial Intelligence workloads now run outside the cloud, and 88% of respondents are running at least one Artificial Intelligence workload on premises. To support that shift, Dell and NVIDIA highlighted NVIDIA Confidential Computing as the basis for protecting models and sensitive data in use while enabling private deployment of frontier systems. Google Distributed Cloud with Gemini 3.0 is now available in preview on Dell PowerEdge XE9780 servers, accelerated by NVIDIA Blackwell and secured by NVIDIA Confidential Computing. SpaceXAI models are also planned for on-premises deployment within the Dell Artificial Intelligence Factory.
The software and ecosystem strategy extended from proprietary and open models to agents and workflow tools. NVIDIA Nemotron models, Reflection open source models and additional open models including MiniMax-M2.7, DeepSeek Pro, DeepSeek-V4, GLM 5.1 and Kimi K2.6 are being made available for enterprise use. OpenAI Codex is set to connect with the Dell Artificial Intelligence Data Platform, with Dell and OpenAI also exploring integration with the Dell Artificial Intelligence Factory. Dell added partnerships spanning Palantir, ServiceNow, Fogsphere, Ipsotek, Mistral Artificial Intelligence, Poolside, Uneeq, CrowdStrike and Fortanix, alongside support for NVIDIA OpenShell, NVIDIA Agent Toolkit, NVIDIA NeMoClaw and the NVIDIA AI-Q Blueprint.
Customer references underscored the production focus. Dell said 5,000 enterprises like Lilly, Samsung and Honeywell are running Artificial Intelligence workloads on Dell Artificial Intelligence Factories with NVIDIA. Lilly discussed life sciences innovation powered by Artificial Intelligence infrastructure deployed at scale, Samsung highlighted chip design and manufacturing use cases, and Honeywell described its move from public cloud to on-premises Artificial Intelligence for industrial use cases, digital twins and automation. Hudson River Trading is also expanding its deployment with Dell PowerEdge XE9685L servers and NVIDIA networking to support Artificial Intelligence-driven research.
