Sunday, October 11, 2026
News

GPU cloud Lambda is raising $4 billion ahead of a 2027 IPO, WSJ reports

Its order backlog jumped from $15 billion in June to $50 billion in September, mostly because of one $35 billion deal with Anthropic.

A close-up of a graphics card on a yellow backgroundAndrey Matveev / Unsplash
Photo: Andrey Matveev / Unsplash

Lambda, which rents out GPU servers as a cloud service, is raising $4 billion in what is expected to be its last private round before an initial public offering, The Wall Street Journal reported on October 7. Blackstone and Coatue Management are leading the round, which values Lambda at $14.5 billion before the new money. Lambda declined to comment to the Journal.

According to a letter to investors seen by the WSJ, Lambda is aiming to go public in 2027.

A backlog that tripled in three months

The same letter put Lambda's backlog of contracted business at $50 billion in September, up from $15 billion in June. Most of the jump came from a deal with Anthropic valued at $35 billion. Microsoft is also a known customer.

Lambda has raised debt as well as equity this year:

Date Financing
August 2026 $926 million senior secured loan facility
Early October 2026 $1 billion in senior secured fixed-rate financing
October 2026 (reported) $4 billion equity round at a $14.5 billion pre-money valuation

Founded in 2012, Lambda runs 15 data centers in the US, according to its website, and is targeting gigawatts of compute capacity by 2030.

What it means for buyers

GPU clouds such as Lambda, CoreWeave and Nebius are now a regular part of AI infrastructure buying alongside AWS, Microsoft Azure and Google Cloud. Lambda's backlog shows how much of that business now comes from AI model companies signing very large contracts. Smaller customers are buying from the same pool of new capacity.

What IT teams should do

  1. If you rent GPUs from Lambda or are evaluating it, ask how much new capacity is already committed to large contracts and what that means for your delivery dates.
  2. For multi-year GPU commitments with any specialist cloud, check what your contract says if the provider is acquired.

Sources