Acquisition memory is another very important topic. Typically it is specified in points, kilo points or mega points. It is memory per channel, plus in some oscilloscopes users can interleave channels to double up on the memory. Memory size defines acquisition time. For example, with a 1 GS/s sample rate and 20,000 points of memory, with a little bit of math the result is a 20 µs capture window. There is a memory sequencing mode that allows the user to optimize the memory for capturing very fast signals where is a need to digitize rapidly and that have large signal gaps or long dwell times between the events. This allows the user to not waste memory, for example if they were capturing a 20 ns pulse every two seconds.