Bernstein analyst argues Core Scientific's 75% return on AI infrastructure is an outlier, not a replicable model for bitcoin miners.
AI & Agents ·
Bernstein research argues that Core Scientific's reported 75% return on investment for AI infrastructure should not be viewed as a blueprint for other bitcoin miners pursuing similar deals. The analyst firm estimates that TeraWulf and Cipher, two other major players in the sector, are achieving stabilized returns on assets of 5% and 4% respectively—substantially lower than Core Scientific's figure. According to Bernstein's analysis, Core Scientific's outsized returns reflect capital expenditure advantages rather than a broadly applicable strategy.
The divergence in returns highlights varying competitive positions across bitcoin mining firms as they diversify into AI infrastructure. Core Scientific's deal structure appears to have provided favorable terms unavailable to other operators, positioning it as an exception rather than an industry standard. The analysis raises questions about whether miners without similar capital advantages can realistically match such returns when pivoting to high-margin AI workloads.
What remains unclear is whether other miners might negotiate comparable deals in future transactions, or whether Core Scientific's position as a larger incumbent provided unique leverage in structuring its arrangement.