Automated Instruction Timing Test
Original Publication Date: 1980-Jun-01
Included in the Prior Art Database: 2005-Feb-13
In the past, timing measurements were made by monitoring some signal which would become active at the beginning of each instruction (e.g., Instruction Fetch), or by doing an address compare on the first and last microinstruction of that instruction's microcode. The CPU was then put into a loop, and the time was measured as the time between pulses. This would obviously be different for each system, and great care had to be taken to insure that the proper signals were chosen to make accurate measurements.