Kim Gunnelius

In the era of generative AI, escalating energy demands of data centers have become a pressing concern. Training AI models is highly energy intensive. According to MIT’s Lincoln Laboratory, the GPUs that trained ChatGPT’s precursor, GPT-3, are estimated to have consumed 1,300 megawatt-hours of electricity – nearly the amount of electricity consumed by 1,450 average U.S. households per month. Yet, amidst this escalating demand, a significant opportunity remains untapped—the recovery and repurposing of waste heat generated by data centers.

The world is at a critical juncture in its battle against climate change. As the consequences of carbon emissions become increasingly apparent, the urgency to transition toward sustainable energy sources has never been more pronounced.

In the ever-changing dynamic landscape of high-intensity compute, Verne Global has always strived to stay one step ahead when it comes to innovation and progress. We evaluate the latest advancements in power, cooling, security and connectivity to make sure our customers can grow, scale and optimise their AI workloads as their data center needs evolve. With this we are pleased to announce the integration of liquid cooling technology on our data center campus, The Rock.