Cloud service providers increasingly prefer high core-count CPUs, thus reducing the number of processors and servers they deploy.