As competition in artificial intelligence (AI) intensifies, so does the demand for the computing capacity and data centres it requires. Citing limits in meeting that demand, Google has restricted the social media giant Meta’s use of its Gemini AI model.
According to a Financial Times report, Google told Meta in March this year that it could not supply the amount of computing capacity or storage Meta wanted, delaying work on several of Meta’s internal AI projects.
Alphabet-owned Google runs the Gemini AI model through its cloud infrastructure. The report said several of Google’s other customers had also faced capacity constraints, but because demand to use Gemini AI was so high, Meta was the most affected.
In response, Meta has directed its staff to use AI resources more efficiently, with particular emphasis on cutting the use of “tokens,” the unit that measures AI-model usage.
Neither Google nor Meta immediately commented on the matter, Reuters reported.
Analysts say the episode shows that ensuring adequate computing infrastructure to keep pace with the rapid spread of AI has become a major challenge for technology firms. Despite billions of dollars invested in advanced chips and data centres, infrastructure constraints have not yet been fully overcome.
Meanwhile, Google Cloud’s revenue rose to $20 billion in the first quarter, to March this year. Alphabet chief executive Sundar Pichai had earlier said, however, that computing-capacity constraints were hampering faster expansion of Google Cloud and that the backlog of customer work had grown.
বাংলায় মূল প্রতিবেদন পড়ুন · Read the original Bengali report