Gigawatt-scale AI data centers are on an accelerated timeline, with many expected to be operational in two years or less. Recent data indicates that construction times for these facilities, which require substantial permitting and infrastructure, can range from 1 to 3.6 years. Notably, xAI has ambitious plans for its Colossus 2, projecting a build time of just 12 months. The first of these large-scale data centers is anticipated to go online by early 2026. The analysis relies on satellite imagery, permits, and company disclosures to track the progress of major U.S. data centers, including those from Amazon and Microsoft. Key milestones include the start of construction, the first operational power draw, and reaching the 1 GW power threshold. The process involves estimating power capacity based on various data points, although the timelines can be approximate due to inconsistencies in reporting and the nature of permit approvals across different jurisdictions. Limitations exist in the data collection, particularly regarding satellite imagery quality and the variability in cooling capacity estimates. The precision of construction start dates can be affected by the availability of high-resolution imagery. Furthermore, companies do not uniformly define what constitutes "operational," which complicates comparisons between facilities. This nuanced approach highlights the rapid growth and complexity of the AI data center sector.