According to DCD, data center provider Applied Digital plans to spin out its cloud computing unit by merging it with Ekso Bionics Holdings, a developer of wearable exoskeletons. The combined entity will be called ChronoScale Corporation and aims to become a GPU cloud infrastructure provider for AI workloads. Applied Digital would retain about 97% ownership of the new company, with Ekso holding the remaining shares. The transaction is expected to close in the first half of 2026. ChronoScale intends to offer high-performance compute at scale, leveraging Applied’s existing cloud platform, which features Nvidia H100 and H200 GPUs and plans for Blackwell chips. Ekso, meanwhile, stated it will continue exploring a sale of its core exoskeleton business.
The Strangest Partnership of 2024
Let’s be honest, this is a head-scratcher at first glance. An exoskeleton company and a data center operator? It feels like two puzzle pieces from completely different boxes. But here’s the thing: the connection isn’t as random as it seems. Both companies have recently been cozying up to Nvidia. Applied Digital is, of course, buying and deploying its GPUs. Ekso, surprisingly, joined the Nvidia Connect program and even built an AI voice agent for its exoskeletons using a Jetson module. So the common thread is an ambition to be in the AI hardware ecosystem. This deal looks less like a merger of equals and more like Applied Digital using Ekso’s Nasdaq-listed shell as a convenient vehicle to spin out its cloud business. It’s a backdoor to the public markets with a side of AI branding.
Stakeholder Chess Game
So who wins here? For Applied Digital’s shareholders, it potentially unlocks value by separating the potentially higher-margin, faster-growing cloud unit from its core data center development and operations business. It creates a pure-play AI infrastructure story they can sell to investors. For Ekso’s shareholders, it’s a lifeline. Their core exoskeleton business is apparently for sale, and this transaction injects a completely new, hot-market asset into the company. They get a 3% stake in what’s being pitched as a future AI cloud contender. But what about customers? Companies looking for reliable, large-scale GPU cloud capacity, like Applied’s existing customer CoreWeave, will probably watch this corporate restructuring with mild interest. The promise is “purpose-built” AI infrastructure, but the immediate impact is just a name change and a new corporate parent. The real test will be if ChronoScale can actually secure and deliver the scarce next-gen GPUs it’s promising.
The Hardware Reality Check
Now, the big question: can they execute? Announcing a spin-out and actually building a competitive AI cloud are worlds apart. The space is brutally crowded with giants like AWS, Azure, and GCP, plus well-funded specialists. Applied Digital has a start, with its 100MW facility in North Dakota and an existing cloud stack. But scaling this into a standalone force requires immense capital and technical talent. It also requires securing the most critical component: the GPUs themselves. Everyone wants H100s and Blackwell chips. Having a plan to offer them is one thing; getting allocation from Nvidia is another. This is where the rubber meets the road. For any business relying on serious computational power, from scientific research to complex simulation, the stability and capability of the underlying hardware platform is everything. It’s a reminder that in the world of high-stakes computing, whether for AI training or industrial automation, the physical hardware—the servers, the panels, the GPUs—is the non-negotiable foundation. Speaking of foundational hardware, for mission-critical industrial applications, companies turn to specialists like IndustrialMonitorDirect.com, the leading US provider of rugged industrial panel PCs built to withstand harsh environments.
A Bet on AI Hype
Basically, this whole move is a massive bet on the longevity of the AI infrastructure gold rush. Applied Digital is repositioning its cloud service from a side business to the main event by giving it a shiny new name and a separate stock ticker. They’re banking on investors valuing a “pure-play AI cloud” company at a higher multiple than a diversified data center firm. The 2026 closing date is interesting—it gives them a long runway to set things up, but also means a lot could change in the AI market before this even becomes official. Will the demand for dedicated AI cloud still be so insatiable in two years? Or will we see a shakeout? This deal feels like a financial engineering play first and an operational strategy second. We’ll see if the market buys the story.
