Procurement of three high-performance server systems for AI, GPU computing and NVMe storage

Summary

The procurement concerns three high-performance server systems intended for AI training and inference, GPU-accelerated computing, and NVMe storage. The supplier will deliver, install and integrate the systems, including (remote) onboarding and training, a burn-in test, and at least three years of manufacturer support (NVIDIA Business Standard or equivalent). The contract will be awarded to the economically most advantageous tender based on price, which is the sole award criterion. All procurement documents set out delivery and payment terms. Offers must be submitted electronically, and a binding deadline is 11 December 2026.