Dell unveiled PowerEdge XE8812 server supporting up to 144 NVIDIA Vera Rubin NVL4 GPUs per rack, establishing new compute density standard for AI infrastructure.
Dell Technologies has announced the PowerEdge XE8812, a new high-performance platform built on NVIDIA Vera Rubin NVL4 architecture designed to support large-scale AI training, inference, engineering simulations, and scientific computing workloads. The system supports up to 144 GPUs per rack and represents an expansion of Dell AI Factory deployments across sovereign AI initiatives, engineering and design workflows, and genomic research.
The fanless, direct liquid-cooled PowerEdge XE8812 combines NVIDIA Vera Rubin NVL4 architecture with NVIDIA CUDA-X libraries to allow organizations to run larger AI models and simulations entirely in memory. The platform delivers 50% more memory per socket and higher GPU memory capacity compared to previous generations, enabling larger AI models and simulations to remain fully in memory and reducing the need for staging data from storage or minimizing swapping operations that introduce latency.
Built on the Open Rack Version 3 (ORv3) standard, the PowerEdge XE8812 is designed for modular deployment and operational efficiency. The liquid-cooled design supports high compute density while improving cooling and power efficiency. The system is configured as a production-ready, factory-integrated and pre-validated rack-scale solution. Combined with Dell ProDeploy services, the platform is designed to reduce deployment complexity and shorten time to production—Dell reports that production-ready racks can be deployed and running workloads in just over six hours.
Dell says more than 5,000 customers worldwide have already deployed Dell AI Factory solutions across AI, engineering, sovereign AI, and scientific workloads. Current deployments include collaboration with Lawrence Berkeley National Laboratory on the Doudna supercomputer, which supports AI training, inference, HPC simulations, and data-intensive scientific research. AI company InstaDeep is expanding its Kyber supercomputing cluster using Dell AI Factory infrastructure, with the cluster expected to deliver approximately 0.5 exaFLOPs of FP16 performance. The Wellcome Sanger Institute is using Dell PowerEdge XE-Series servers with NVIDIA GPUs for large-scale genomic research, while Monash University, in collaboration with Dell, NVIDIA, and CDC Data Centres, has deployed the MAVERIC supercomputer to support AI and data-intensive research in cancer detection, climate action, and genomics.
The PowerEdge XE8812 will be available globally in early 2027. Venkat Sitaram, Senior Director and Country Head, Infrastructure Solutions Group, Dell Technologies India, stated: "As India advances its AI ambitions and organizations increasingly seek to leverage AI and high-performance computing to drive innovation, the need for scalable, energy-efficient and future-ready infrastructure has never been greater. With the Dell AI Factory with NVIDIA and the introduction of next-generation platforms such as the Dell PowerEdge XE8812, we are enabling enterprises, research institutions and emerging sovereign AI initiatives with the performance, density and open architecture required to address demanding AI and HPC workloads. Dell remains committed to helping customers in India translate AI ambition into real-world outcomes at scale."
Chris Marriott, Vice President, Enterprise Platforms, NVIDIA, added: "The convergence of AI and HPC is reshaping what organizations can expect from their infrastructure. Dell and NVIDIA are advancing that benchmark together, combining NVIDIA Vera Rubin NVL4 architecture and CUDA-X libraries with Dell's engineering capabilities and large-scale deployment expertise to deliver the performance, efficiency and openness needed for some of the world's most demanding AI and scientific computing workloads."