Crypto miners are meeting the artificial intelligence (AI) industry’s requirements by transforming their data centers in response to the increase in GPU demand.
As per a recent report by Bloomberg, crypto mining companies are identifying new opportunities in the AI sector by converting their existing infrastructures, as AI has an unprecedented demand for data centers and GPU resources.
The AI Industry is the Primary Driver of the Demand for Computing Power
The growth of AI has resulted in an unprecedented scarcity of industry-powering equipment and supplies, as well as the electricity necessary to power these resources.
According to a March report by commercial real estate firm CBRE Group, AI companies and cloud service providers have already claimed approximately 83% of the capacity of new facilities currently under construction.
Many companies are utilizing crypto-mining firms to satisfy the increasing demand for data center space. These firms already have the requisite infrastructure and are in search of more profitable applications for their properties.
At a conference hosted by Nvidia, Core Scientific Chief Operation Officer Matt Brown recounted his experience as “just getting cold-called, or getting text messages from industry peers that are looking for capacity because they are finding that it is just not there.”
According to Brown, “Everyone is frantically trying to determine where they can locate capacity within the next 12 months, and the task is becoming exponentially more challenging.”
The Partnership between Core Scientific and CoreWeave
This trend of collaboration was underscored by Core Scientific’s announcement of a partnership with AI startup CoreWeave, which is anticipated to generate approximately $3.5 billion in revenue over 12 years.
Core Scientific CEO Adam Sullivan stated, “We perceive the current AI opportunity as an opportunity to convert our existing infrastructure to host clients interested in installing vast arrays of GPUs for their clients, who are ultimately AI clients.”
The company has identified approximately 500 megawatts of capacity that is available for conversion to AI facilities, which has the potential to be the most extensive GPU installment dedicated to AI globally. This capacity adequately supplies electricity to hundreds of thousands of residences annually.