Michelle Weaver, an American subject research strategist at Morgan Stanley, said a few days ago that corporate applications of AI are being implemented at an accelerated pace, but limited computing power supply is still a bottleneck limiting the growth of the industry. The current supply of computing power is seriously inadequate. Computing power is becoming a scarce resource,” Weaver said in an interview with the media on Wednesday. Weaver pointed out that there are two main reasons for the computing power bottleneck: one is the lack of labor required to build the data center, and the other is the lack of electricity to power the data center. Artificial intelligence data centers consume huge amounts of electricity, and it takes years to build new power generation facilities, transmission networks, and related infrastructure. Even considering innovative power supply solutions such as the transformation and utilization of Bitcoin mining facilities and fuel cells, Weaver estimates that there is still a power gap of 10% to 20%, which means that computing power will still be limited and a high-value resource for the next few years. This means that even if companies have sufficient capital to buy AI chips and build data centers, insufficient electricity or labor may hinder the actual launch of computing power. In addition to insufficient electricity and labor, political factors are also one of the obstacles to the expansion of AI computing power supply. Weaver also pointed out that as the midterm elections approach, rising anti-data center sentiment is challenging the industry.

Michelle Weaver, an American subject research strategist at Morgan Stanley, said a few days ago that corporate applications of AI are being implemented at an accelerated pace, but limited computing power supply is still a bottleneck limiting the growth of the industry. The current supply of computing power is seriously inadequate. Computing power is becoming a scarce resource,” Weaver said in an interview with the media on Wednesday. Weaver pointed out that there are two main reasons for the computing power bottleneck: one is the lack of labor required to build the data center, and the other is the lack of electricity to power the data center. Artificial intelligence data centers consume huge amounts of electricity, and it takes years to build new power generation facilities, transmission networks, and related infrastructure. Even considering innovative power supply solutions such as the transformation and utilization of Bitcoin mining facilities and fuel cells, Weaver estimates that there is still a power gap of 10% to 20%, which means that computing power will still be limited and a high-value resource for the next few years. This means that even if companies have sufficient capital to buy AI chips and build data centers, insufficient electricity or labor may hinder the actual launch of computing power. In addition to insufficient electricity and labor, political factors are also one of the obstacles to the expansion of AI computing power supply. Weaver also pointed out that as the midterm elections approach, rising anti-data center sentiment is challenging the industry.