Oxide Computer has closed a $445 million Series D funding round led by Eclipse, valuing the private cloud infrastructure company at $6 billion.

The funding—bringing Oxide's total raised to approximately $835 million—reflects a widening appetite among enterprises to escape the per-compute economics of hyperscalers. Oxide's core pitch is straightforward: give companies the software and hardware stack to run cloud-grade infrastructure on-premises, eliminating the margin compression that comes with renting compute from Amazon Web Services, Microsoft Azure or Google Cloud.

The math is brutal for large-scale users. Hyperscalers operate on gross margins north of 60 percent on compute services. An enterprise running consistent workloads can build owned infrastructure, amortize it over five to seven years, and cut per-unit costs by 40 to 50 percent. The catch is operational complexity and engineering overhead—precisely what Oxide is selling its way past.

Oxide bundles custom-designed hardware (optimized for specific workload classes) with its homegrown operating system and orchestration software. The model mimics how hyperscalers themselves operate internally: vertically integrated stacks that eliminate handoff costs. For Oxide, unit economics improve as customers deploy larger footprints and sticky software lock-in deepens.

The $6 billion valuation sits on an unproven revenue base. Oxide operates in a market segment—private cloud infrastructure for Fortune 500 engineering teams—that generates enormous contract values but remains thin in public investor eyes. Hyperscale vendors have spent a decade and billions in R&D to commoditize this problem. Oxide's bet is that the economic pendulum has swung far enough that a pure-play alternative can capture share from customers who view cloud-as-utility economics as unsustainable.