History of IBM CKD Controllers
Beginning with its 1964 System/360 announcement, IBM's mainframes initially accessed count key data (CKD) subsystems via a channel connected to separate Storage Control Units (SCUs) with attached direct-access storage devices (DASD), typically hard disk drives. This practice continued in IBM's larger mainframes thru IBM Z; however low end systems generally used lower cost integrated attachments where the function of the SCU was combined with that of the channel, typically called an Integrated File Adapter.
Source: Wikipedia — History of IBM CKD Controllers (CC BY-SA 4.0)