The Charged Coupled Device (or CCD) was initially conceived in 1970 at Bell Labs by W. Boyle and G. Smith. These researchers were interested in producing an electronic analog to the commonly used bubble memory in order to improve computer mass storage capability. Stored charged readout along the registers of the CCD would then allow this device to be a serial memory device. The first imaging CCD was produced by Fairchild Electronics in 1974 and had a format of 100x100 pixels. The charge transfer efficiency of this device was so low, that the overall yield was less than 0.5% (slightly less than good photographic plates at the time). Needless to say, at this time, observational astronomers were somewhat less than enthusiastic about replacing glass plates with digital electronics. But, by 1979, this had all changed as an RCA 320x512 LN$_2$ cooled CCD system would see first light on a 1-meter telescope at Kitt Peak National Observatory.
Early observations with this CCD quickly showed its superiority over photographic plates. The quantum efficiency was at least a factor of 50 higher (in the red), the device itself was highly linear (photographic plates are highly non-linear) and thus easy to calibrate, and the response of the detector was fairly uniform thus facilitating the detection of faint, low contrast nebulosity. Over the last 15 years, the CCD has been steadily improved and is evolving towards an ideal detector. Such a detector would have 100% Quantum Efficiency, perfectly uniform response, be noiseless, have unlimited dynamic range, have unlimited numbers of pixels and would have completely understandable characteristics. Todays current generation of multi-phase pin, thinned, back illuminated 2048x2048 devices comes close to reaching those goals.