Microsoft released its new Tabular model of Analysis Services almost three years ago. Even if it is relatively new for many developers, we already used it in several production scenarios. In this session, we will share some of the hard lessons learned from the first large deployments in Tabular. How do you handle billions of rows with complex calculations? What tools do you use to benchmark and choose the right hardware? Why throwing money on the hardware could be a bad idea? How do you scale up performance on large databases? How do you make a very large database fit in memory? This session will be of most interest to BI developers, especially those who are considering or starting to implement an analytical solution in the tabular model.