Iris Energy, listed on Nasdaq, has purchased 248 state-of-the-art Nvidia H100 GPUs for $10 million as it seeks to explore opportunities in productive artificial intelligence alongside its main focus on Bitcoin mining. Nvidia supports the company’s growth in this field by introducing the GH200 Grace Hopper Superchip to meet Iris Energy’s artificial intelligence requirements.
The company expects the delivery of the 248 GPUs in the coming months and plans to use the hardware to serve opportunities in cloud computing. Daniel Roberts, Co-founder and Co-CEO of Iris Energy, stated that the company aims to utilize its existing data centers to meet the productive artificial intelligence computing needs.
“We believe that the demand for sustainable computing will continue, and whether it’s ASICs for Bitcoin mining or GPUs for general artificial intelligence and beyond, we believe we are in a unique position to capture ongoing growth in the broader sector.”
Iris Energy operates in regions abundant in renewable energy such as wind, solar, and hydroelectric power, and establishes modular data centers near low-cost excess renewable energy sources that can be converted into money for Bitcoin.
According to the Iris Energy website, the company has four major data center mining facilities, including Canal Flats, Mackenzie, and Prince George in British Columbia, Canada, as well as the Childress facility in Texas. Renewable energy-powered Bitcoin mining operations continue to attract investments with Genesis Digital Assets Limited opening a new data center in Sweden in August 2023 that will operate with ample power from the nearby Porjus Hydroelectric Power Plant.
Meanwhile, Blockstream recently announced its intention to raise up to $50 million in a formal investment note to acquire, store, and later sell BTC mining hardware before the next halving event in 2024. Nvidia, the GPU hardware manufacturer, has also seen significant growth from the rise of AI-powered tools and AI computations, with its total market value surpassing $1 trillion in May 2023.
Nvidia has also recently introduced the next-generation GH200 Grace Hopper Superchip, which is said to handle complex productive AI workloads, including large language models, recommendation systems, and vector databases.