Dell has launched new high-end servers featuring top Nvidia GPUs for enterprise AI workloads, bundled in an ‘AI Factory’ package with software, networking, and managed support – all aimed at accelerating AI adoption and driving sales.
In sum – what to know
New servers – a bunch of new high-end Dell servers integrate top-line Nvidia GPUs for enterprise cloud and edge AI workloads.
Factory package – hardware is being bundled with new software and networking switches as part of a joint ‘AI Factory’ pitch.
Drive and sell – the masterplan (!) is to accelerate enterprise adoption of AI, and sell lots of servers and software licences and such.
Dell and Nvidia have deepened their AI dealings with the introduction of new AI server hardware, integrated AI software, and fully managed AI services – all geared to enterprises, deploying AI across edge and cloud setups. The move seeks to make high-end AI accessible to enterprises that lack skills, infrastructure, and capacity to build and run domain-specific AI engines themselves. The plan is to accelerate enterprise adoption of AI, and sell lots of servers and software licences and such.
Sounds obvious, but the pair want to be the go-to edge and hybrid cloud/edge AI infrastructure providers for enterprises, and not just for big centralised cloud hyperscalers in the cloud. At Dell Tech World in Las Vegas, they launched a bunch of new offerings as part of their joint ‘AI Factory’ (Dell AI Factory with Nvidia – in full) initiative. These combine Dell’s infrastructure with Nvidia’s GPUs and AI software. In sum, Dell is launching a raft of new servers, all available in the second half of the year, for accelerated compute and data processing, integrated with new network switches, an updated management platform, and 24/7 managed support.
The new servers include air-cooled PowerEdge XE9780 and XE9785 units (for “integration into existing enterprise data centers”) and direct-to-chip liquid-cooled PowerEdge XE9780L and XE9785L units (pictured; to “accelerate rack-scale deployment”). These are the successors to the PowerEdge XE9680, presented as Dell’s “fastest ramping solution ever”. They support up to 192 Blackwell Ultra GPUs (“customized” to 256) per Dell’s highest-end server cabinets – specifically its IR7000 system. The Ultra series, introduced at GTC 2025, sits at the top of Nvidia’s GPU portfolio
Dell reckons the new units deliver up to four times faster large language model (LLM) training with an eight-way Nvidia HGX B300.4 configuration – where eight Blackwell Ultra GPUs are interconnected via fifth-generation NVLink, the chip firm’s latest GPU-to-GPU interconnect technology, part of the Blackwell architecture. There are other servers besides: the PowerEdge XE9712, featuring Nvidia’s GB300 NVL72 system, offers “50-times more AI reasoning inference output and and five-times improvement in throughput”; and the PowerEdge XE7740 and XE7745, available with Nvidia’s enterprise-geared RTX Pro 6000 GPUs in July, and supported in Nvidia’s ‘Enterprise AI Factory’ design.
This XE7740/XE7745 platform works as a “universal platform to help meet the needs of physical and agentic AI use cases like robotics, digital twins, and multi-modal AI applications”, said Dell. It supports up to eight GPUs in a 4U (4×1.75 inch rack unit) chassis. Dell said it will provide a new PowerEdge XE server for Nvidia’s upcoming Vera Rubin supercomputing architecture platform, an advance on its Blackwell series, geared for ultra-high-density compute and advanced memory and networking technologies.
Meanwhile, Dell is offering support for Nvidia’s upgraded Ethernet and InfiniBand networking switches (PowerSwitch SN5600, SN2201 Ethernet, Quantum-X800) to deliver up to 800 gigabits per second of throughput to match the demands of AI workloads on the new servers. Besides, Dell is offering “enhancements” to its AI Data Platform, specifically related to its ObjectScale and PowerScale storage platforms, and with integration with Nvidia’s BlueField-3 DPU network card and Spectrum-4 Ethernet switch chip. Introduction of KV cache and RDMA support, now provide lower-latency access to data during inference, it said.
There’s other stuff too: Dell is offering Nvidia’s AI Enterprise software suite, including microservices, retrieval models, reasoning models, tools for developing agentic AI models and workflows, plus support for Red Hat OpenShift. Dell is offering a managed service for the whole AI Factory solution, taking care of everything in the Nvidia AI stack with 24/7 monitoring, reporting, version upgrades, and patching.
The new Dell Managed Services for the Dell AI Factory with Nvidia simplify AI operations with management of the full Nvidia AI solutions stack — including AI platforms, infrastructure and Nvidia AI Enterprise software. Dell managed services experts handle 24×7 monitoring, reporting, version upgrades and patching, helping teams overcome resource and expertise constraints by providing cost-effective, scalable and proactive IT support.
Michael Dell, chairman and chief executive officer at Dell Technologies, said: “We’re on a mission to bring AI to millions of customers around the world. Our job is to make AI more accessible. With the Dell AI Factory with Nvidia, enterprises can manage the entire AI lifecycle across use cases, from training to deployment, at any scale.
Jensen Huang, founder and chief executive officer at Nvidia, said: “AI factories are the infrastructure of modern industry, generating intelligence to power work across healthcare, finance and manufacturing. With Dell, we’re offering the broadest line of Blackwell AI systems to serve AI factories in clouds, enterprises and at the edge.”