Intelligent CIO LATAM Issue 02 | Page 42

FEATURE : DATA CENTERS
Hyperscale sites reach Chile , Colombia and Mexico
Hyperscale data centers operated on the colocation model host the largest cloud providers on the planet . The demand for space in these data centers is such that , in some cases , it requires the construction of huge works of approximately 5,000 square meters .

IT SHOULD BE NOTED THAT A COLOCATION SITE CONSIDERED SMALL TODAY IS FOUR TIMES THE SIZE OF WHAT A SMALL COLOCATION DATA CENTER WAS 10 YEARS AGO .

The rate of expansion of this type of work is so rapid that some suppliers are remodeling their internal processes to fully perform as a customercentric company . In this model , all internal teams – engineering , projects , logistics , services , finance and human resources – have a single focus : anticipating and solving the current and future needs of colocation data centers , an ever-evolving environment .
ISPs and telecom operators are also competing for this market
In recent years , colocation data centers of various sizes , implemented by telecommunications providers and ISPs ( Internet Service Providers ) are also multiplying . It should be noted that a colocation site considered small today is four times the size of what a small colocation data center was 10 years ago .
The growing adoption of Edge Computing solutions in Latin America will contribute to the number of colocation data centers increasing even further . According to the study The Data Center 2025 : Closer to the Edge of the Grid , conducted in 2019 by Vertiv and based on interviews with 800 data center managers , the number of computing sites in the Edge will grow by 226 % by 2025 ( global data ).
According to AFCOM , the American Association of Data Center Professionals , these are large , hyperscale data centers . Mega sites are up to 22,000 square meters . In Latin America , the expansion movement of hyperscale colocation data centers began in Brazil and , throughout 2020 , progressed towards Chile , Colombia and Argentina , and continues with the construction of units in other Latin American countries , such as Mexico .
With the development of 5G networks and the growth in demand for data processing close to the point of data consumption , this expansion will be even greater . By the end of this decade , thousands of these centers are expected to exist in Latin America .
São Paulo , Rio de Janeiro , Buenos Aires and Mexico City are the regional centers
Many data centers are owned by large companies that want to ensure the consistency of their infrastructure throughout Latin America . These service providers are certified as Tier IV or Tier III .
Uptime Institute certifications are the standard for the global data center project , construction and operation market . Tier levels are tailored to business needs , with higher levels required by the infrastructure to provide greater redundancy and fault tolerance . Maintaining these seals requires excellent and on-going management , regardless of where the hyperscale colocation data center is located .
The demand for homogeneous and high availability processes by the Latin American data center industry is leading the large colocation sites in the region to require a regional , and not a local , perspective from their suppliers in relation to their critical infrastructure .
In terms of geography , the expansion of colocation data centers can be seen in all regions . The presence of these data centers has been a reality in Brazil for several years – 65 sites , according to Data Center Map . Recently , countries like Peru , Uruguay and Paraguay are also working on colocation projects .
An Equinix study , published in April 2020 , shows that in Latin American cities , 90 % of the data processed by their colocation data centers in the region passes through São Paulo , Rio de Janeiro , Buenos Aires and the City of Mexico .
The expansion of colocation data centers in Latin America continues to face challenges . The competitiveness of these companies depends more and more on the optimization of energy consumption . A study conducted by the US government in 2016 showed that data centers
42 INTELLIGENTCIO LATAM www . intelligentcio . com