Lambda Raises Up to $3bn Before IPO, Months After Borrowing $917m for Chips
Nvidia-backed GPU cloud, Lambda, is seeking up to $3bn in pre-IPO financing at a valuation of $12bn or more, with a listing planned for next year, according to Bloomberg.
The company, which rents out Nvidia accelerators and surrounding infrastructure to train and serve AI models, follows a different financing path earlier this year when it borrowed $917mn against its GPUs to acquire more chips from Nvidia.
The Business Model
Lambda’s business is straightforward: providing GPU cloud services to companies training and serving AI models. They compete with hyperscalers on price and with a growing field of specialist neoclouds on availability.
Benefits of Pre-IPO Financing
Equity financing ahead of an IPO offers advantages over leveraged loans. Debt secured against depreciating hardware is risky, and a public market debut is easier to price when the balance sheet is not heavily weighed down by chip-related borrowings.
The Nvidia Connection
Nvidia’s involvement in Lambda is noteworthy. They are an investor, supplier, and indirectly a source of demand, a strategy replicated across the sector, with Nvidia’s equity commitments exceeding $40bn this year.
Investors have not considered this circularity a disqualifier, assuming the demand is genuine. However, questions arise about the extent of this demand.
Customer Traction and Future Concerns
Lambda has secured a cloud agreement with Hudson River Trading to provide access to Nvidia chips, demonstrating real customer traction. Yet, the long-term sustainability of neoclouds remains uncertain as customers build their own silicon or negotiate directly with hyperscalers.
Timing an IPO is delicate, especially with market volatility and lenders paying closer attention to funding sources. Contract structure is crucial; long-term commitments from creditworthy customers provide stability, while spot rentals expose startups to funding market fluctuations.
Additionally, rising input costs, particularly Nvidia server price increases above 15% due to memory shortages, impact Lambda’s cost base.
Public markets also bring enhanced scrutiny, particularly regarding depreciation schedules, which directly affect reported profitability.