In 1965 I had a problem and access to a real time DN-30 computer that was running the exec for the Dartmouth time sharing system. So I added some tasks to calculate some results on the DN-30. I was using a good fraction of the computer time and the blinkenlights showed an unusual pattern as a result. I got my answers and turned off the program and then realized I could get the same answers, by hand, with paper and pencil quite quickly.
Since the computer was a real time computer there wasn't any appreciable slowdown.