Life is too short to wait for simulation's results.
We need useful and accurate results and we want it now.
How much time you can wait? For human driven simulations, few milliseconds create the illusion of istantaneus response. For real time simulation-driven control, the response time can decrease to nanoseconds. Solid state electronic circuits are entering in the vacuum tubes age. Moore's law is finished. They simply cannot continue to scale up as in the past. Electronic circuits needs too much energy, too much materials and too much space.
To continue the scale-up process we need a different approach to the problem. To solve large scale probes we need to look to the bottom. Actually, there is plenty of space in the bottom.
Working in the sub nano meter scale is a very particular job.
As start, forget classical determinism. 1+1 = 2 sometime, if you average the results.
But it is important? Hard determinism produces better results? Try to shot ten rounds to a target 1000 meters away. If you are very good, you can put it all inside a 0.1 meter diameter circle. Tango down. Absolute accuracy is not required.
The very same simulation produces two slightly different results if you run it twice with the very same inputs?
This is not a sign of malfunction but natural intelligence.
