發生了什麼事
Nscale, the London‑based AI infrastructure firm, filed an S‑1 registration statement with the SEC on September 18 and applied to list on the New York Stock Exchange under the ticker “NSCL.” Lead underwriters will be Goldman Sachs, JPMorgan and Morgan Stanley. The filing does not yet specify the offering size or price range, and the registration statement has not become effective, leaving the final listing date subject to regulatory review and market conditions.
Nscale is an AI infrastructure company headquartered in London, operating data centers, power assets, GPU clusters and cloud software. It spun out in 2024 from Australian cryptocurrency infrastructure firm Arkon Energy and quickly pivoted from Bitcoin mining to AI compute services. The firm now employs roughly 450 staff and is one of Europe’s larger AI‑infrastructure operators by funding raised.
At the end of August, Nscale reported about 7 MW of capacity in its own data centers and an additional 48 MW in leased facilities, representing contract values of roughly $2.6 billion. The company has about 1.3 GW of projects in planning or construction, with future revenue hinging on the timely activation of this capacity.
In March 2024, Nscale closed a $2 billion financing round at a $14.6 billion valuation, backed by Nvidia, Dell and Nokia. Dealroom data shows total equity financing of $3.7 billion and over $5 billion in debt. If the IPO reaches the $35 billion target cited by the Financial Times, the market value would have more than doubled in six months.
為什麼這很重要
The filing signals a major step for an AI‑focused cloud provider to tap US capital markets, potentially raising billions to fund a rapid expansion of GPU‑heavy data centers across Europe and the United States. A valuation target of up to $35 billion would more than double the $14.6 billion valuation from a March $2 billion funding round, reflecting investor confidence in the growing demand for AI training and . Nscale’s close partnership with Nvidia—both as a core chip supplier and a multi‑billion‑dollar investor—adds credibility and may accelerate GPU adoption among European AI firms, while also exposing the company to concentration risk in the Nvidia ecosystem.
Nscale’s business model—leasing large‑scale GPU on long‑term contracts to AI developers, tech firms and governments—addresses a critical bottleneck as AI models grow in size and demand rises. By providing dedicated AI‑optimized infrastructure, the firm helps customers avoid the capital expense of building their own data centers.
The partnership with Nvidia goes beyond chip supply. Nvidia has invested more than $2 billion in Nscale, signed lease agreements worth about $1.2 billion, and guaranteed $860 million of lease obligations for a Texas facility. This deep integration gives Nscale preferential access to Nvidia GPUs and may steer AI workloads toward Nvidia’s ecosystem, reinforcing Nvidia’s market position in Europe.
However, the model is capital‑intensive. Building power, cooling and infrastructure requires massive upfront spending, while revenue is recognized only after capacity is delivered. Delays in construction, power supply constraints, or shifts in customer compute demand could increase costs and defer cash flow, a risk reflected in the company’s first‑half revenue of roughly $141 million versus a near‑$1 billion net loss.
互動機制：它實際上是如何運作的
以互動方式探索這項發展背後的基礎技術。
As use of AI scales up across an organization, what tends to matter most?
接下來看什麼
Investors and industry observers will monitor Nscale’s ability to bring its 1.3 GW of planned capacity online on schedule, the terms and timing of its US IPO pricing, and how its long‑term lease contracts with customers such as Anthropic translate into cash flow. The company’s debt load, the progress of its West Virginia super‑ campus, and broader AI‑sector sentiment—especially around safety concerns raised by leading AI CEOs—will also influence the success of the offering.
The timing and pricing of the IPO will reveal how much capital Nscale can raise and at what valuation, influencing its ability to fund the remaining 1.3 GW of planned capacity.
Progress on the flagship West Virginia data center, which includes a six‑year, up‑to‑$45 billion lease with Anthropic, will be a key indicator of future revenue visibility.
Broader AI‑sector sentiment, especially regulatory scrutiny and safety concerns voiced by leaders at OpenAI, Anthropic and other firms, could affect demand for new capacity and thus impact Nscale’s growth trajectory.