Institutional Bitcoin Mining Investors Mostly Care About AI’s Data Center Potential — Not BTC, Bernstein Says

Institutional investors are being drawn to bitcoin mining companies for their artificial intelligence (AI), but they may reap unexpected benefits from cryptocurrency bull markets.

“If we're right about our $200K Bitcoin price prediction, investors will come to AI and enjoy Bitcoin bull markets without asking,” Bernstein's report released Monday. He might not have expected it.

This unexpected convergence between AI infrastructure and cryptocurrency mining is attracting attention from institutional investors who are concerned about the traditional boundaries of these technologies.

According to Bernstein, Bitcoin miners are uniquely placed in this position due to their high power resources and strategic location.

“Collectively, Bitcoin miners have access to 4GW of power today, with an operational network that will move to 6GW connections by the end of 2024,” he said, pointing to the massive infrastructure these companies have at their disposal.

On the issue of re-rating for Bitcoin miners, Bernstein analysts write that they have a comparable power portfolio, yet trade at a discount to legacy data centers. ~$4mn/MW to ~$30-50mn/MW, but miners earn $-$0.6mn/MW versus $4.7mn/MW.

“As Bitcoin miners operate on AI data center campuses, we believe this difference in both revenue and business multiples will decrease,” Bernstein said.

According to Bernstein's analysis, Bitcoin miners have a high value in terms of energy infrastructure.

These operations currently have a capacity of 4GW, which is projected to reach 12GW by 2027.

This massive power capacity, coupled with miners' experience running high-density operations at 70-80KW per rack, matches the power requirements of AI computing.

Unlike traditional data centers, Bitcoin miners have strategically built their operations around “fragmented energy sources,” often in unusual locations where land and energy are abundant.

This approach has led to the creation of sprawling sites covering hundreds of hectares with power capacities ranging from 100MW to 1GW.

For example, the Terra Wolf (WULF) Lake Marine Station in Western New York has 500 MW of hydropower capacity, enough to cool water sources – a critical factor for both crypto mining and AI operations.

Bernstein also noted that Bitcoin miners' expertise extends beyond access to power and their profitability hinges on sophisticated energy cost management, including hedging strategies in wholesale markets and cooperative relationships with utility and grid operators.

This knowledge can be useful in managing the power-intensive needs of AI computing.

Edited by Stacy Elliott.