Core Insights - The order book for the company stands at approximately $450 billion, which is expected to translate into revenue over the next several years as data centers become operational [1] - The company is positioned in the AI cloud space by renting out GPUs for model training and inference workloads, similar to other hyperscalers [2][6] - The company has a high-margin software and applications business that supports its expansion in cloud infrastructure, unlike competitors who may need to borrow at higher costs [3] - By the end of the decade, the company anticipates a potential upside of $60 billion compared to consensus estimates [4] - The company has recently upgraded its infrastructure and is sourcing chips from Nvidia to enhance its offerings [5] - There is a significant shortage of computing capacity in the market, leading to increased demand for the company's services [6][7] - The company's cloud infrastructure revenue was around $10 billion last year, significantly lower than competitors like Amazon and Azure, which reported $125 billion and $75 billion respectively [8] - The company has been undergoing a transformation since 2017, shifting focus towards providing computing capacity and services [9]