正文 · AI 翻译
IT Home reported on October 27 that, according to Reuters, Bank of America, Blackstone, Citi and Morgan Stanley will provide $63 billion in financing (IT Home note: approximately 420.679 billion yuan at the current exchange rate) to purchase Broadcom artificial intelligence chips and networking equipment for Anthropic. It is one of the largest debt transactions related to the construction of AI computing capacity.
As demand for artificial intelligence continues to rise, major technology companies are racing to secure computing capacity to support increasingly complex workloads, driving massive investment in computing infrastructure.
The companies involved said on Monday local time that the transaction will be carried out through a newly established special-purpose vehicle, which will purchase Broadcom AI accelerator chips and networking equipment and then lease them to Anthropic, the developer of Claude.
Bank of America, Citi and Morgan Stanley are arranging the first-lien financing, while Blackstone is providing the committed funding portion of the second-lien financing.
Broadcom is providing “residual value support” to the senior debt lenders, effectively giving lenders a guarantee under the financing structure: Broadcom is willing to cover losses arising from a decline in the assets’ value, up to an amount equal to the first-lien debt. The announcement, however, did not disclose the size of that debt tranche.
The companies said the equipment is expected to begin delivery in 2027 and is part of a previously announced strategic agreement to supply computing capacity in which Broadcom is involved.
Earlier, Apollo and Blackstone had already reached an agreement with Broadcom to provide $35 billion (approximately 233.711 billion yuan at the current exchange rate) to expand Anthropic’s AI computing capacity.
The latest transaction underscores the increasingly important role banks and alternative asset managers play in financing infrastructure to meet growing demand for AI computing capacity.
Anthropic filed for an IPO on Monday local time, and the offering could set a record. At the same time, disclosures of heavy losses and safety risks could test investors’ willingness to invest in the AI giant.