Around March, Google, a subsidiary of Alphabet, informed Meta that it was unable to fulfill the complete Gemini capacity that Meta aimed to purchase, according to the article. This limitation has disrupted and delayed some of Meta’s internal AI initiatives.
Other clients of Google have also experienced impacts, albeit to a lesser degree. However, Meta has faced significant challenges due to its exceptionally high demand for Google’s models, as noted by the FT.
Reuters was unable to confirm the report immediately, which referenced sources familiar with the situation. Neither Google nor Meta has responded to requests for comment outside of business hours.
As a result of these limitations, Meta has advised its employees to use AI tokens, the units that quantify AI usage, more efficiently, according to the FT report.
Despite continued investments of billions in chips and data centers, companies are still finding it difficult to obtain sufficient computing power to meet the rising demand for AI services.
In the first quarter ending in March, Google Cloud’s revenue increased to $20 billion. However, CEO Sundar Pichai indicated that constraints on computing power hindered even greater growth and led to a nearly doubling of the cloud unit’s backlog quarter over quarter.
