Queretaro, a central Mexican city, has emerged as the country's data-centre capital with scores of warehouse-like server facilities operated by major firms including Microsoft, Amazon Web Services and ODATA. Ascenty maintains two roughly 20,000 sq ft sites in the state with a third under construction. Forecasts estimate more than $10bn in data-centre investment over the next decade. AI demand is accelerating data-centre construction. Queretaro's central location, proximity to Mexico City, high-speed cable connectivity and supportive municipal and federal permitting and zoning make it attractive. US power-grid capacity limits are driving some firms to seek power in locations like Queretaro.
Across the state companies including Microsoft, Amazon Web Services and ODATA own these warehouse-like buildings, full of computer servers. No one could supply an exact number, but there are scores of them, with more being built. Ascenty, which claims to be the largest data centre company in Latin America, has two in Queretaro, both around 20,000 sq ft in size, with a third under construction.
It is forecast that more than $10bn (7.4bn) in data centre-related investment will pour into the state in the next decade. "The demand for AI is accelerating the construction of data centres at an unprecedented speed," says Shaolei Ren, associate professor of electrical and computer engineering at the University of California Riverside. So, what's the attraction of Queretaro? "It's a very strategic region," explains Arturo Bravo, Mexico country manager at Ascenty.
"Queretaro is right in the middle [of the country], connecting east, west, north and south," he says. That means it is relatively close to Mexico City. It is also connected to high-speed data cables, so large amounts of data can be shifted quickly. Mr Bravo also points out that there is support from the municipality and central government. "It's been identified as a technology hub," he says. "Both provide a lot of good alternatives in terms of permits, regulation and zoning."
Collection
[
|
...
]