CoreWeave Buys Core Scientific in $9B Data Center Deal
▼ Summary
– CoreWeave signed a $9 billion all-stock deal to acquire Core Scientific, a data center infrastructure provider.
– The acquisition gives CoreWeave access to over a gigawatt of data center capacity, which can be rented for AI workloads.
– Core Scientific previously offered Bitcoin mining but will now focus on running and training generative AI models.
– Cloud providers are expanding data centers to meet the growing computational demands of AI companies.
– OpenAI recently secured a deal to rent an additional 4.5 gigawatts of data center capacity from Oracle.
The AI infrastructure race intensifies as CoreWeave makes a massive $9 billion acquisition of data center specialist Core Scientific. This all-stock transaction positions CoreWeave to significantly expand its computing capacity, securing over one gigawatt of data center power – equivalent to the electricity needs of more than 850,000 households.
This strategic move reflects the surging demand for AI computation resources, with companies scrambling to secure infrastructure capable of handling complex machine learning workloads. Core Scientific, originally known for Bitcoin mining operations, will now pivot its powerful GPU clusters toward training and running generative AI models under CoreWeave’s ownership.
The acquisition comes amid fierce competition in cloud infrastructure. Recent reports reveal OpenAI’s agreement with Oracle to lease an additional 4.5 gigawatts of data center capacity, building upon their existing Stargate partnership. These developments highlight how AI’s exponential growth is reshaping data center markets, with providers expanding rapidly to meet unprecedented computational needs.
By absorbing Core Scientific’s assets, CoreWeave strengthens its position as a key player in AI-optimized cloud services. The deal underscores how specialized computing infrastructure has become a critical battleground as organizations vie for dominance in the booming artificial intelligence sector.
(Source: TechCrunch)
