1. They restore and restart the simulation from the last known stable snapshot. So you don't know about a downtime even if one happened. 2. While testing out a

1. They restore and restart the simulation from the last known stable snapshot. So you don't know about a downtime even if one happened. 2. While testing out a RTC program on 8085 processors in college we used to run the clock faster just to check if Clock counted to all conditions correctly. There is a always a fast forward option to jump the states you are interested in. 3. For the people running the simulation the time between big bang and heat death could be something like 5 mins what we perceive as billions of years.