Singapore’s SuperX AI Technology just made its first move into North America, choosing Denver of all places for its debut AI inference cloud location on the continent. It’s a notable pick. Most companies chasing AI cloud headlines tend to gravitate toward the usual suspects, Virginia, Texas, or somewhere overseas, so landing in Colorado says something about where cloud capacity is actually available right now.
Details remain thin so far. SuperX hasn’t said how much capacity its Denver cloud site offers, who’s operating the underlying data center, or exactly which hardware configuration is running there, beyond confirming it relies on Nvidia accelerators. The company’s website does list Nvidia’s B200 and B300 chips, along with GB300-based rack systems and its own SuperX XI6150 server, so the new deployment likely draws from that same lineup.
What stands out more than the hardware specifics is the demand SuperX claims to have already lined up. According to the company, unnamed AI-focused companies reserved a “significant portion” of the cloud site’s inference capacity even before launch. SuperX is currently running access through an invitation-only onboarding process, which suggests the company is starting small and selective rather than opening its doors wide from the start.
“The AI industry is entering a new phase where success is increasingly defined by the ability to operate AI services reliably and efficiently at scale,” said Kenny Sng, the company’s CTO, framing the launch around growing demand for cloud infrastructure that targets inference workloads specifically rather than general-purpose computing.
This Denver launch fits into a broader pattern of cloud expansion for SuperX. The company partnered with STT GDC on a Singapore facility earlier this year, signed a memorandum with several partners in Japan to explore a data center there, and even launched a joint venture last year focused on high-voltage power systems for AI infrastructure. Denver itself already hosts cloud and data center operations run by companies like Equinix, Flexential, and Iron Mountain, so SuperX is entering a cloud market that’s far from empty.
Whether the city becomes a genuine hub for AI inference cloud work, or just one stop in a much wider global expansion, probably depends on how SuperX scales beyond its current invite-only phase.
