Thanks to time-dependent simulations such as Large Eddy Simulations but also modelling more complex industrial flows, demands on HPC resources in CFD have increased dramatically in recent years. Typical time-dependent simulation uses large number of CPUs during long time (from weeks to months). Engineers in industry are simulating more complex flows and more care is put on details (such as representing all details in the geometry). All this would not be possible without cost efficient HPC resources in form of Linux clusters. This minisymposium will give an overview of HPC-intensive CFD applications covering computation of flows in water turbines, flows around cars, trains and airplanes and radiation heat transfer.