In November 2024, CoreWeave and Core Scientific announced the commencement of construction on a new 100-megawatt (MW) data center located at the Port of Muskogee in Oklahoma. Scheduled to become operational in 2026, this facility is designed to support high-performance computing tasks, particularly those involving artificial intelligence (AI) and machine learning (ML) applications.
The data center will house NVIDIA GPUs provided by CoreWeave, catering to the computational needs of an undisclosed client. Of the total capacity, approximately 70MW will be dedicated to critical IT workloads, while the remaining 30MW will support ancillary systems essential for optimal GPU performance. This project exemplifies the collaborative efforts between CoreWeave and Core Scientific to advance AI infrastructure. Adam Sullivan, CEO of Core Scientific, highlighted the significance of this development, emphasizing Muskogee’s emerging role as a hub for next-generation AI computing.
The initiative received substantial support from local authorities, including the Port Muskogee team and Mayor Patrick Cale, as well as the state of Oklahoma, underscoring the region’s commitment to technological advancement and economic growth.
This data center represents a portion of the 500MW capacity that CoreWeave has contracted from Core Scientific for high-performance computing hosting operations. In addition to constructing new facilities, Core Scientific is repurposing some of its existing cryptocurrency mining sites into AI and high-performance computing hosting centers, relocating Bitcoin mining hardware to other locations as part of this transition.
The Muskogee project not only enhances the region’s technological infrastructure but also signifies a strategic shift towards supporting the growing demands of AI and ML workloads. By leveraging their combined expertise, CoreWeave and Core Scientific aim to establish a state-of-the-art facility that will contribute significantly to advancements in AI computing capabilities.
This development highlights the dynamic evolution of data center operations to accommodate the increasing computational requirements of emerging technologies.