Lambda is seeking to raise as much as $3 billion prior to its initial public offering, just a few months after securing $917 million in loans for semiconductor production.

      According to Bloomberg, Lambda is in discussions to raise up to $3 billion in pre-IPO funding at a valuation of $12 billion or higher, as the Nvidia-backed cloud provider prepares for a public listing anticipated next year. This financing initiative contrasts sharply with an earlier phase in its operations when Lambda secured $917 million by leveraging its GPUs to acquire more chips from Nvidia, which is also an investor in the company.

      The business model is easy to articulate but costly to maintain. Lambda leases Nvidia accelerators and associated infrastructure to companies that train and deploy AI models, positioning itself against hyperscalers on pricing and an increasingly competitive landscape of specialist neoclouds concerning availability.

      Equity financing ahead of the IPO offers advantages that the leveraged loan could not provide. Debt secured by depreciating hardware is a rigorous tool, and entering the public market becomes simpler when the balance sheet does not heavily rely on borrowings tied to chips with unpredictable resale values.

      Negotiations are not finalized, and discussions remain private, so the list of investors and final funding amount have not been disclosed. Lambda has not shared any public comments on the matter.

      The dynamics of Lambda's relationship with Nvidia merit attention. Nvidia acts as an investor, supplier, and indirectly fuels demand; this arrangement has been mirrored throughout the sector, leading Nvidia's equity commitments to surpass $40 billion this year. Most companies receiving these funds allocate a significant portion toward Nvidia hardware, which boosts demand in a manner that the debt markets have started to factor in. Lambda exemplifies this structure rather than being an exception.

      To date, investors have not viewed this circularity as disqualifying, based on the reasonable assumption that a supplier taking equity in its customers poses an issue only if the underlying demand is not genuine. However, there is ongoing debate regarding the actual extent of this demand.

      Customer traction appears solid. For instance, Lambda has entered a cloud agreement with Hudson River Trading to provide access to Nvidia chips—a partnership type that conducts thorough diligence regarding uptime and costs.

      A broader concern for any neocloud is what unfolds once the capacity crunch subsides. These entities generate their margins from scarcity, and the same clients currently renting GPUs may pivot to developing their own silicon or negotiating directly with hyperscalers in the coming years.

      Timing an IPO under such conditions is sensitive. Lambda would be entering a market that has reevaluated AI infrastructure multiple times, where lenders have become markedly more vigilant about who is truly financing expansion.

      The structure of contracts is crucial for the survival of these companies. Long-term agreements with reputable customers lead to cheaper debt and a straightforward equity narrative, whereas short-term rentals to startups leave a fleet of high-cost hardware vulnerable to the fluctuations of future funding markets.

      Additionally, rising input costs complicate the financial picture. Nvidia has informed customers of AI server price hikes exceeding 15% due to memory shortages, which directly impacts the cost basis for a company that operates by renting out those servers.

      Public markets also impose scrutiny that private rounds do not, especially concerning depreciation schedules. The speed at which a company depreciates GPUs influences its reported profitability, and there is no established standard across the sector regarding the expected lifespan of an accelerator's earnings.

      A $12 billion valuation would place Lambda beneath the largest neoclouds while firmly within the bracket of companies investors anticipate to enter public markets. This listing has been expected since at least last year when reports first indicated the company was preparing for it.

      The competitive landscape has also evolved, with CoreWeave already public and several competitors awaiting their turn. What once appeared as a unique listing is now one among several, with comparable trading multiples for investors to consider.

      No aspects of this situation are finalized. The funding round is still under negotiation, the valuation remains a goal rather than a definitive figure, and an IPO planned for next year may well shift to one scheduled for the following year.