Hi SSC, I have a Production AWS Windows Server with 8 cores, SQL 2016 Enterprise installed, MaxDOP setting of 8 at Instance/Server Level, MaxDOP of 0 at Database Level, CostThresholdForParallelism of 35. My understanding is that every query estimated to take 35 or more seconds to execute will go parallel, using all 8 cores (leaving none available for other processes). As a result other processes will have to wait (blocking) until at least one or more cores become available based on its estimated cost of execution.