1. What is Nscale asking the public market to pay?
Nscale, the London-based AI infrastructure company, is seeking a stock market valuation of up to $35 billion in its planned New York Stock Exchange listing, and reports on October 18 say some prospective investors are resisting that price. The company filed a Form S-1 registration statement to list under the ticker NSCL, and it aims to raise about $3 billion from new shares.
The pushback arrives at an awkward moment for companies that build computing capacity for artificial intelligence. Enthusiasm for the sector has cooled in recent weeks, construction timetables at several sites have slipped, and at least one comparable company has dropped its own listing plans. Nscale's pricing ambition therefore faces a more doubtful audience than the one that welcomed earlier offerings in the segment.
The headline number deserves some care. Accounts of the target differ: most coverage of the filing cites a ceiling of $35 billion, while some reports put the figure closer to $30 billion. The final range will only be known once the company publishes the number of shares and an indicative price, which had not been set when the registration statement was filed.
2. How does the target compare with Nscale's earlier valuation?
Nscale's most recent private valuation was $14.6 billion, set in a Series C funding round in March 2026. A $35 billion listing value would be roughly 2.4 times that figure, reached in about seven months. Even the lower $30 billion figure would more than double the March mark.
Investors in a public offering usually test such a jump against growth and profitability. On growth, the numbers are striking. Revenue for the six months to June 30 reached $140.6 million, compared with $10.4 million in the same period a year earlier, an increase of 1,252%. A business that expands from a small base at that pace can justify a premium multiple, at least in the eyes of its supporters.
Profitability tells a different story. The filing shows a net loss of $1.02 billion, several times larger than the revenue booked over the same stretch. That is typical of infrastructure builders, which must pay for land, power connections, chips and construction long before customers generate meaningful income. It also means the valuation rests on expectations about future revenue rather than present earnings, which is exactly where skeptical buyers tend to press.
3. Why does the contract backlog not settle the argument?
The strongest argument in Nscale's favor is its contracted business. The company reports roughly $103.4 billion in total contracted customer value, up from about $38 billion at the end of 2025. A substantial portion of that comes from agreements signed with Anthropic on August 25, which account for around $44.6 billion. Demand on that scale explains why management feels able to aim high.
The weakness lies in how little of that backlog is producing revenue today. As of late August, only about 2.5% of the contracted value was active, meaning that the overwhelming majority of the money depends on data centers that still have to be completed, powered and handed over to customers. A contract only becomes income when the capacity behind it is switched on.
Concentration is a related concern. Most of the contracted demand comes from a small number of customers, so a change of plans by any one of them could reshape the revenue outlook. For a prospective shareholder, a large backlog with few counterparties is valuable but also fragile, and that trade-off is a natural basis for demanding a lower entry price.
4. What is slowing the buildout?
Execution is the second area of doubt. Nscale's Norwegian facility is running about six months behind schedule. It has 6.4 megawatts of capacity operating against a target of 30 megawatts, which is around 21% of the planned total.
Such delays matter more for a young company than for an established one. Data center capacity earns nothing until it is energized, while financing costs and construction commitments keep accumulating. Delays in bringing sites online therefore push back revenue without postponing the expenses attached to the project.
For the IPO, the practical consequence is that investors want evidence that the 2.5% activation rate will climb quickly. Progress on the Norwegian site and on other projects will be read as a test of whether management can convert its signed contracts into running capacity, and that evidence will influence how much buyers are prepared to pay.
5. Why is the market for AI infrastructure listings so difficult now?
Nscale is not marketing its shares in isolation. Several offerings by AI infrastructure companies have been postponed or canceled, and Australia's Firmus Grid abandoned its listing plans entirely after investors pushed back. When a peer withdraws, remaining issuers inherit a more demanding set of buyers.
Listed competitors are not offering comfort either. CoreWeave, the best-known publicly traded company of this kind, saw its shares fall about 15% in a single week. The group is often called neoclouds, a term for specialist providers that rent graphics processing power to customers running AI workloads. They operate in a narrower lane than giants such as Amazon or Microsoft, and they usually depend on borrowed money to build their capacity.
That dependence on debt is why a falling share price in one company quickly spreads to its peers. Investors who doubt that today's heavy spending will be repaid by tomorrow's customer payments tend to apply that doubt across the whole group, and a new issuer must persuade them that it is the exception.
6. What did the pre-IPO convertible notes do?
Before the public offering, Nscale already secured a large block of private funding. On September 25, 2026, it closed a pre-IPO convertible note round of $3.36 billion, led by hedge fund Third Point. These notes convert into equity at a discount to the eventual IPO price, a feature that rewards the early lenders for taking risk ahead of the listing.
The notes also carry a cap tied to a $30 billion valuation. Above that threshold, the conversion price stops adjusting. That detail makes $30 billion a reference point for the whole offering: if the IPO prices at or below it, the discount mechanism fully protects the noteholders, and new public shareholders effectively absorb the dilution that follows when the notes convert.
The gap between the $30 billion reference and the $35 billion ceiling is one reason the valuation debate is closely watched. Buyers in the public offering will weigh whether they are being asked to pay more than sophisticated private lenders effectively agreed to. A price closer to the lower figure would narrow that gap and might ease the resistance.
7. What should investors watch before Nscale starts trading?
The next milestone is the disclosure of the price range and share count in an updated filing, which will show how far the company moves from its initial ambition. A range centered on $30 billion would suggest that management has heard the feedback, while a range near $35 billion would signal that it expects strong demand to override the objections.
Beyond pricing, operational updates will carry weight. Evidence that the Norwegian site is catching up with its schedule, and that a rising share of the $103.4 billion backlog is becoming active, would give buyers firmer ground. Any further delay or change of customer commitments would do the opposite.
Finally, the broader sector will set the tone. The outcome of other AI infrastructure offerings, and the share price behavior of CoreWeave in the coming weeks, will shape how receptive the market is. The company's listing date has not been announced, and the timing could still shift if conditions deteriorate.
8. How do convertible notes and IPO pricing interact?
A convertible note is a loan that can turn into shares instead of being repaid in cash. Lenders accept a lower interest burden, or none, because they receive the right to own equity at a favorable price once the company lists. In Nscale's case, the favorable price is set as a discount to the IPO price, so the lower the offering is priced, the less each note converts into, up to the stated cap.
This structure links the interests of early and late investors in a way that is easy to overlook. A higher IPO price helps the company raise more cash per share sold, yet it also raises the price at which the notes convert, which benefits existing shareholders. A lower price, by contrast, makes the offering easier to sell but shifts more ownership toward noteholders. Bankers running the book must balance these effects while gauging demand.
Market participants also treat convertible rounds as a signal of confidence. Experienced investors committing billions before a listing indicates belief in the business, yet the discount they negotiated shows they wanted compensation for the risk. Public buyers will weigh both messages when deciding whether the share price asked of them is fair.
9. Why do valuation targets matter beyond one company?
A listing of this size acts as a reference for an entire category. If Nscale prices near its ambition and trades well, other builders of AI computing capacity gain a template for raising money publicly. If it must cut its target sharply, or delays the offering, the result would reinforce the caution already visible in the shelved plans of others.
The outcome also bears on how the market judges the large commitments being made by AI developers. Contracts such as those signed with Anthropic only become reliable income streams if the providers can complete and power their sites. A well-received offering would suggest that investors accept the timeline, whereas a poorly received one would suggest that they want proof before paying for growth that has not yet happened.
For now, the situation is a negotiation between a company confident in its contracted demand and buyers focused on losses, delays and a hesitant sector. Both sides have credible arguments, and the final price will show which one carries more weight.