Data centers are an integral part of the modern era. In the current connected world, there is an ever-increasing demand for data storage and by 2025, half of the world’s data will reside in the cloud. This demand has las led to a significant growth in the size and number of available data centers in recent years. The forecast by Gartner shows that spending on data centers grew from $194 billion in 2020 to $226 billion in 2022, and by the end of 2021, there are projected to be more than 600 hyperscale data centers worldwide. These data centers must be efficient and robust, handling equipment failure, network congestion, communication or power outages and thermal management at peak loads. Managing cooling through Heating, Ventilation and Air Conditioning (HVAC) systems and rack layout is one of the most important factors in meeting these challenges in data center design. In this study, we will show how simulation can help to improve cooling effectiveness, thus improving data center performance and reducing operating costs. The HVAC System model is coupled with a detailed 3D Computational Fluid Dynamics (CFD) model representing the server room. The 3D CFD model includes multiple rack clusters, sensors and other components to capture the details of this complex multiphysics and multiscale problem. The co-simulation underlines issues with cold and hot aisle separation, as well as hot spots due to air recirculation, and helps to select optimal sensor positioning. In addition to this, the co-simulation approach allows predicting the time to shutdown of the server equipment under the effects of an HVAC component failure. This integrated simulation can be carried out in early design stages to assess and optimize the cooling needs for any type of data center.