IREN has entered a purchase agreement to buy Nvidia Blackwell AI systems from Dell Technologies for about $1.6 billion [1, 2, 3, 4, 5].
This investment represents a significant scaling of physical infrastructure to accommodate the global surge in artificial intelligence workloads. By securing high-end hardware, the company aims to maintain a competitive edge in the AI cloud business and meet soaring demand for specialized computing power [2, 3, 4].
The deal focuses on Nvidia's air-cooled Blackwell systems [1]. These units will be installed at IREN's Childress campus in Texas, U.S. [1, 5]. According to industry reports, the commissioning of these systems is expected to occur in early 2027 [5].
Market reaction to the announcement was immediate. IREN shares saw a four percent rise in pre-market trading following the news [3]. The agreement was announced on May 26, 2026 [4], and reported by various financial outlets on May 27, 2026 [3].
The collaboration between the data center operator and the technology giant underscores the ongoing race to build out the physical layer of the AI economy. The deployment of Blackwell systems allows for greater efficiency and processing power compared to previous generations of hardware.
IREN continues to expand its footprint in the U.S. to support the infrastructure requirements of AI cloud services [1, 5]. The move aligns with a broader trend of data center operators diversifying their hardware portfolios to include the latest GPU architectures to attract high-scale AI clients.
“IREN entered a purchase agreement to buy Nvidia’s air‑cooled Blackwell AI systems from Dell for about $1.6 billion”
This deal signals a critical transition for IREN as it pivots more aggressively toward AI cloud services. By committing $1.6 billion to the latest Blackwell architecture, the company is betting on the long-term demand for massive-scale compute. The choice of air-cooled systems suggests a specific infrastructure strategy for the Texas climate and site requirements, aiming for a balance between performance and operational scalability.
