Dell boosts enterprise AI hardware with AMD partnership in Ireland and Nvidia‑based XE8812 launch

Dell Technologies and AMD announced a joint effort to expand on‑premises AI capabilities for Irish enterprises. Starting in July 2026, Dell PowerEdge servers will be equipped with AMD Instinct MI350P PCIe GPUs, delivering up to 4,600 teraflops and 144 GB of HBM3e memory. The solution enables generative and agentic AI workloads without requiring data‑centre redesign and integrates AMD’s software stack, including ROCm and the Enterprise AI Suite, into the Dell AI Platform.

At the same time Dell unveiled the PowerEdge XE8812, a new AI‑factory server built around NVIDIA’s Vera Rubin NVL4 architecture. The rack‑scale system supports up to 144 GPUs per rack, features direct liquid cooling, expanded host memory and higher GPU memory capacity, and is aimed at large‑scale AI training, inference, HPC simulations and scientific research. Dell cited deployments with partners in the United States, France and the United Kingdom, and highlighted rapid, pre‑validated rack deployment that can be operational within hours.