NVIDIA has named Rubin as the generation after Blackwell. Pacific is designing the next configuration of the pod around it, and the work is under way now, alongside Pod 32.
NVIDIA DSX POD is NVIDIA's reference design for containerized AI factories on the Vera Rubin platform, built by partners and delivered as complete modules. NVIDIA says a DSX POD can be running about 14 days after it reaches a prepared site.
Pacific's Vera Rubin configuration is being designed to the DSX POD reference architecture. The intent is the same 40 foot container, with the power and cooling inside reworked for the new platform, so the site, the power connection and the way the pod is operated carry over from Pod 32 to the next generation.
The intent is the same container, with the power and cooling inside reworked for the new platform. A data center design now has to carry more than one generation of GPU, and the aim is that the site and the connection stay put while the compute inside moves on.
The Vera Rubin configuration is in development and there is no specification to publish yet. Pod 32, with NVIDIA B300s, is the current configuration. If you are planning around Rubin, book a call.
Pacific rents GPUs, builds managed clusters and sells pods to selected partners.