NVIDIA partners with IREN, Span, and PulteGroup to deploy AI infrastructure across utility-scale facilities and residential settings

NVIDIA and IREN Limited agreed to deploy up to 5 gigawatts of next generation AI infrastructure. NVIDIA also partnered with Span and PulteGroup to create distributed fractional data centers that use residential grid capacity for AI workloads.

These partnerships expand NVIDIA's reach from large-scale facilities into distributed, community-level compute nodes—extending its GPU and software platforms beyond traditional data centers into utility-scale sites and residential-linked infrastructure.

For investors, the deals link NVIDIA more tightly to how and where AI compute is built, and signal how AI infrastructure projects may allocate capital, power capacity, and hardware across a wider mix of models going forward.