Steven Glagola

Steven Glagola asked to clarify if Bitdeer's U.S. AI cloud expansion, particularly in Washington and Tennessee, is contingent on securing multi-year reserve capacity agreements, and if these commitments would primarily be for bare metal deployments. He also sought Jihan Wu and Matt Kong's perspective on the strategic attractiveness of U.S. AI cloud expansion and Bitdeer's long-term competitive advantages beyond its Asia-centric footprint.

Answer

Haris Basit, Chief Strategy Officer, confirmed that any significant large-scale GPU expansion in the U.S. is dependent on signing customer contracts, though small speculative expansions are possible. Jihan Wu, Chief Executive Officer, highlighted owning data centers in the U.S. as a crucial advantage, enabling deployment of X300 and routing by year-end. He emphasized the U.S. as a global AI innovation hub with strong demand, where customers prefer capacity on U.S. soil, positioning Bitdeer favorably.

Steven Glagola sought clarification on whether the new deal's margin is comparable to Core42 after accounting for corporate SG&A and asked about key learnings from the initial data center builds that will improve future efficiency.

Answer

CFO Patrick Fleury reiterated that the 85% margin is project-level NOI, which excludes corporate SG&A. CTO Nazar Khan and COO Sean Farrell highlighted key learnings, including the momentum from retaining contractors, experience with diverse GPU designs, and accelerating construction through detailed upfront design collaboration with customers.

Stephen Glagola asked if the high concentration of CoreWeave as a customer could be a deterrent for other hyperscalers and inquired about the level of enterprise demand for the Grand Forks site specifically.

Answer

CEO Adam Sullivan argued the opposite, stating that their execution for CoreWeave, delivering 250 MW in a single year, will breed confidence among other potential customers, including hyperscalers. He added that the Grand Forks site has seen interest from all segments, including enterprises, content providers, and some hyperscalers, for its high-density colocation development potential.