Everpure CFO says hyperscalers are lining up for AI data-center buildout on a spending scale ‘twice’ the U.S. defense budget

Everpure, the data-storage and infrastructure company formerly known as Pure Storage, put a stark number on the economics of the AI buildout at its Accelerate 2026 summit in Las Vegas. In remarks reported by Yahoo Finance, CFO Tarek Robbiati said hyperscale cloud providers are preparing to spend on AI infrastructure at a scale that he characterized as “twice the entire U.S. defense budget.”

While the reported comment did not spell out a methodology, the message aligned with what many large-technology investors watch right now: the race to expand capacity for GPUs, power, cooling, and, crucially, the storage and data systems that feed high-throughput AI training and retrieval workloads. In the same context, the report named major hyperscalers, including Google and Microsoft, alongside Amazon and SpaceX, as participants in the buildout cycle.

The CFO’s framing matters because it links AI spending not just to compute. In practice, AI datacenters require large, fast, and reliable data pipelines, with storage systems designed for rapid access, high bandwidth, and sustained performance as models iterate and workloads scale. For companies like Everpure, the implication is that infrastructure purchases are likely to be measured in multi-year refresh and expansion cycles rather than one-off deployments tied to a single model release.

Robbiati’s remarks also point to a broader market dynamic that has been visible across the sector: hyperscalers are increasingly competing on end-to-end platform capacity. That includes the back-end infrastructure that moves and retains training data, logs, feature sets, and application datasets used for inference, analytics, and operational monitoring. When those systems become bottlenecks, storage and infrastructure vendors can see demand rise even as the conversation often centers on chip supply and model performance.

For Microsoft in particular, AI capacity has become a central theme of its cloud strategy. Microsoft sells AI services through Azure, and it has continued to market AI capabilities that run on its cloud platform. However, the Yahoo Finance report on Everpure’s comments did not attribute specific AI infrastructure spending totals or contract values to Microsoft, nor did it describe any particular procurement program or hardware refresh timeline tied to the company’s remarks.

The report also did not provide additional detail about the named comparison number, such as whether the estimate includes power and facilities spend, networking, compute, software, or only data storage. It similarly did not clarify whether the figure is a forward-looking projection, an annualized run rate, or a cumulative multi-year total. Without that breakdown, the “twice the defense budget” line should be treated as a high-level announcement of intensity rather than a precise budgetable item.

From a sector standpoint, the immediate takeaway is that AI infrastructure is drawing capital from the same cloud operators that already dominate enterprise IT spending, and that the winners are likely to include infrastructure specialists as well as cloud platforms. As hyperscalers expand, demand can shift from pure compute-centric investments to the systems that keep data moving at scale, which is where vendors like Everpure aim to capture share.

What to watch next is whether companies in this ecosystem start translating broad spending intensity into more concrete indicates, such as order visibility, customer implementation timelines, or guidance that connects AI workload growth to measurable revenue drivers. For Everpure, that would likely show up in updates about demand for its storage and data infrastructure offerings, while for hyperscalers like Microsoft, it could show up in disclosures about capacity expansion and AI service scaling that go beyond marketing language.