Glitch is the transient response observed on the output of the DAC when transitioning from one code to another. The maximum glitch typically occurs during a major code transition, for example, from 01111 to 10000, due to the fact that this code transition causes all of the internal switches to change state. During part characterisation, data is now gathered on all successive code transitions for ADI DACs.  Glitch is specified in nanovolt seconds, which describes the area of the transient response. In the case of the AD5060  nanoDAC, this is specified as being less than 3 nanovolt seconds. The peak-to-peak amplitude is also specified, in this case as being less than 3.5 millivolts.  The important point to note here is that the glitch is of a very short duration. While it may be easily observed on an oscilloscope, the end application may well have a much slower response time, making the glitch invisible to the system. Indeed the delay of some systems can be dominated by the PID control algorithms, which can be hundreds of milliseconds in duration.