Dell has introduced the PowerEdge XE8812 server with NVIDIA Vera Rubin NVL4. The system joins the Dell AI Factory with NVIDIA platform, which Dell says has been deployed by more than 5,000 customers.
The new server is aimed at high-performance computing and artificial intelligence workloads. It is designed for use in Dell PowerRack 9100, an Open Compute Project standards-based rack architecture, and can scale to 144 GPUs per rack.
The launch comes as technology groups seek to meet growing demand from governments, research bodies and large companies running larger AI models and more complex simulations. Dell is positioning the XE8812 as part of a broader push to supply infrastructure for those workloads, particularly as AI and traditional supercomputing increasingly overlap.
The server uses a fanless design with direct liquid cooling for both CPUs and GPUs. Dell says the move from NVIDIA GB200 NVL4 to NVIDIA Vera Rubin NVL4 brings more host memory, more cores and more GPU memory than the previous generation.
The rack design is based on the ORv3 open standard and supports more than 300kW of power. The system can be managed through Dell's Integrated Dell Remote Access Controller, Integrated Rack Controller and OpenManage Enterprise tools, which provide monitoring and leak detection across rack-scale deployments.
Deployment time is another part of the pitch. Factory-integrated PowerRack systems are designed to replace manual integration with pre-validated racks that can be deployed and run live workloads in just over six hours.
Customer use
Dell linked the launch to several customer projects across research, engineering and national AI programmes. In the US, Dell, NVIDIA and the National Energy Research Scientific Computing Centre are building Doudna, the next flagship supercomputer for the US Department of Energy at Lawrence Berkeley National Labouratory.
Doudna will use Dell PowerEdge XE8812 servers with NVIDIA Vera Rubin NVL4 and NVIDIA Quantum-X800 InfiniBand networking. Dell says the system is intended to handle larger HPC workloads, AI training and inference, and data-intensive workflows.
In France, Dell and NVIDIA are supporting AI company InstaDeep as it scales its Kyber supercomputing cluster. Dell says the system delivers about 0.5 exaFLOPs of FP16 performance and is used for large-scale AI model training and industrial design work, including the automated design of printed circuit boards.
In the UK, the Wellcome Sanger Institute is using Dell PowerEdge XE-Series servers with NVIDIA GPUs in genomic research. Dell says the institute now produces one fully assembled genome every seven hours and manages more than 100 petabytes of curated genetic data on premises.
Australia is another market highlighted by Dell. Monash University has developed and deployed the MAVERIC supercomputer with Dell, NVIDIA and CDC Data Centres, using liquid-cooled Dell PowerRack systems and Dell PowerEdge XE9712 servers with NVIDIA GB200 NVL72 architecture.
Broader demand
The launch reflects a wider contest among infrastructure suppliers to capture spending on AI training, inference and scientific computing. Vendors are increasingly focusing on denser systems, liquid cooling and integrated rack-scale designs as data centre operators face constraints in power, space and thermal management.
Dell says AI investment is projected to rise 44 per cent year on year in 2026, while 87 per cent of organisations say innovation is key to their business strategy and 75 per cent say the same of AI. Those figures help explain why large hardware providers are tying server launches to broader software, deployment and services offerings.
Dell also used the announcement to highlight the growth of the Dell AI Factory with NVIDIA, which it says now has more than 5,000 customers worldwide. That customer base spans sovereign AI programmes, research institutions and commercial users in sectors including engineering, life sciences and industrial design.
Ahead of the announcement, Dell and NVIDIA framed the launch around rising computational demands in both science and industry.
"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 organizations the density, memory and open architecture they need to tackle workloads that once seemed impossible," said Arun Narayanan, Senior Vice President, Compute and Networking, Dell.
"The convergence of AI and HPC is redefining what organizations 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 Marriott.