.webp)
Dell has launched the PowerEdge XE8812 server with NVIDIA Vera Rubin NVL4 architecture. The system extends Dell's AI Factory with NVIDIA, now used by more than 5,000 customers worldwide.
The server is aimed at organisations running artificial intelligence and high-performance computing workloads, including scientific simulation, model training and data-intensive research. It can scale to 144 GPUs per rack and uses direct liquid cooling across CPUs and GPUs.
The launch comes as technology groups and research institutions invest in infrastructure that can support larger AI models and more complex simulations on the same systems. Dell is positioning the XE8812 as part of that shift, with a rack design based on the Open Compute Project's ORv3 standard and management software including iDRAC, OpenManage Enterprise and the Dell Integrated Rack Controller.
Dell said the server offers 50% more memory per socket than the previous generation, along with increased GPU memory. That is intended to let customers run larger models and simulations in memory rather than moving data in and out of storage, which can slow performance.
Research focus
One of the first major deployments will be Doudna, the next flagship supercomputer at the National Energy Research Scientific Computing Centre in the US. Dell, NVIDIA and NERSC are building the system at Lawrence Berkeley National Labouratory for AI, HPC and data-intensive scientific work across fields from molecular science to astronomy.
Dell also pointed to a broader set of customer projects to highlight demand for AI and supercomputing systems. In France, AI company InstaDeep is scaling its Kyber supercomputing cluster with Dell and NVIDIA technology. In the UK, the Wellcome Sanger Institute is using Dell PowerEdge XE-Series servers with NVIDIA GPUs for large-scale genomic analysis.
Dell also highlighted work in Australia with Monash University, which has developed the MAVERIC supercomputer with Dell, NVIDIA and CDC Data Centres. The system is intended to support research in areas including cancer detection, climate science and genomics.
Infrastructure push
The XE8812 is part of a broader effort by Dell to sell integrated AI infrastructure rather than individual hardware components. Its PowerRack approach is designed to provide pre-integrated rack-scale systems that reduce the work needed to deploy large computing clusters.
According to Dell, the rack can support more than 300kW of power and is fully liquid-cooled. The design is intended to improve deployment efficiency in data centres where power, space and thermal limits increasingly shape buying decisions.
Management and monitoring are another part of the pitch. Dell said the rack-level controller and OpenManage software provide telemetry and leak detection, while iDRAC allows remote deployment, updates and monitoring of PowerEdge servers.
The announcement also reflects the close relationship between Dell and NVIDIA as server makers compete for a growing share of AI infrastructure spending. Much of that market has centred on systems built around NVIDIA chips, with vendors seeking to differentiate through integration, cooling design and service support.
Arun Narayanan, Senior Vice President of Compute and Networking at Dell Technologies, said the target market includes major public research bodies and organisations building national AI systems. "The institutions doing the world's most important research like decoding the human genome, modeling the energy systems of the future and building the sovereign AI infrastructure that nations depend on deserve infrastructure that matches the ambition of their work. The Dell PowerEdge XE8812 reflects Dell's commitment to pushing the boundaries of what's possible, giving these organisations the density, memory and open architecture they need to tackle workloads that once seemed impossible," Narayanan said.
NVIDIA framed the product as part of a broader convergence of AI and traditional supercomputing workloads. "The convergence of AI and HPC is redefining what organisations should expect from their infrastructure. Dell and NVIDIA are raising that bar together, combining NVIDIA Vera Rubin NVL4 architecture and CUDA-X libraries with Dell's engineering and at-scale deployment expertise to provide the performance, efficiency and openness required for the world's most demanding AI and scientific computing workloads," said Chris Marriott, Vice President of Enterprise Platforms at NVIDIA.