Neocloud PaleBlueDot AI Raises $200M Series C at $3.2B Valuation

Money keeps pouring into AI infrastructure. Last month alone, Crusoe raised a $3.9 billion Series F at a $30.9 billion valuation and Nscale secured $3.36 billion in new funding. Now a GPU-cloud startup that crossed the $1 billion mark in January has roughly tripled its valuation in eight months: Silicon Valley’s PaleBlueDot AI.

PaleBlueDot said it has raised $200 million in a Series C led by ComputeCore, at a valuation of $3.2 billion. The company gave no description of ComputeCore. B Capital, the existing shareholder that led January’s $150 million Series B, joined along with other global investors; that round valued the company at “more than $1 billion.” The new money will go toward additional compute capacity, giving customers more choice in locations and hardware.

Three Businesses on One Platform

PaleBlueDot calls itself a “Super Intelligence Infrastructure Platform,” and the business now has three arms: self-owned GPU clusters that it customizes for customers with large or specialized workloads, a GPU marketplace that offers on-demand compute from a global network of supply partners, and a pay-as-you-go serverless inference service for developers and enterprises. At the time of the Series B, the company ran two of those — the marketplace and dedicated clusters — so serverless inference is the new addition.

It also shared some performance figures. Its NVIDIA B300 cluster in Japan recently earned NVIDIA Exemplar Cloud status, meaning its performance was validated against NVIDIA’s reference benchmarks. As of the end of September, the company says it had signed more than $5 billion in customer contracts, and customers in the U.S. and Japan together account for over half of monthly revenue. That $5 billion is contract value as reported by the company, not revenue, and PaleBlueDot has not disclosed its revenue.

Equity isn’t its only source of capital. In July, PaleBlueDot secured a three-year, $255 million private note from Brookfield Asset Management and Tor Investment Management, with JPMorgan as placement agent, to refinance an existing credit facility and expand AI infrastructure for enterprise customers. GPUs are expensive to buy, and many neoclouds mix equity and debt.

The Founder and the CEO

PaleBlueDot was founded in 2024 by Jonathan Zhu together with Shaodong Huang. Zhu studied at Peking University and Stanford Graduate School of Business, and the company had previously raised a $10 million Series A from family offices. In January it brought in enterprise-technology veteran Stephen Watts as CEO, tilting the company toward large customers and longer-term contracts.

“We will continue to broaden our customer base, with a focus on frontier labs, Neolabs and enterprises in the U.S., and we will invest in our full-stack and go-to-market teams,” Watts said. “Neolabs” is industry shorthand for research-first AI startups founded by researchers who came out of big labs such as OpenAI and DeepMind.

PaleBlueDot operates across North America, Japan, South Korea, and Southeast Asia including Singapore. In Korea it has supplied AI compute to enterprise customers.

Competitors: Where Neoclouds, Inference Platforms, and GPU Marketplaces Overlap

PaleBlueDot sits where three different kinds of competitors meet.

The first group is neoclouds with their own clusters. Leader CoreWeave received a $2 billion investment from NVIDIA in January to build out 5 GW of AI factories by 2030. Lambda raised $1.5 billion in November last year, and Crusoe and Nscale both closed major rounds last month.

In serverless inference and open-source model hosting, Together AI raised $800 million in July at an $8.3 billion valuation. Fireworks AI raised $1.5 billion, and Baseten has been pursuing $1.5 billion at a $13 billion valuation. All three go head to head with PaleBlueDot’s serverless inference business.

In GPU marketplaces, RunPod has been raising its profile, including in Korea’s AI ecosystem, after a roughly 150 billion won ($100 million) investment.

Spanning all three, PaleBlueDot’s $3.2 billion valuation sits below specialist inference players like Together AI and Baseten. The question is how fast the $5 billion in contracts it points to turns into actual revenue.