Report this content

You might be able to do something clever. If you wire duplicate logic in series you can get things to process 1 after the other on the same frame. I am not sure if Dreams has some kind of cut off before execution is moved to the next frame? But this could help you achieve faster then light exeuction.

So for example you provide value 1 into a chip. Inside the chip you process something and output that signal. Then duplicate the chip and feed that value 1 into the input of the next. Because of the way signals work the second chip has to wait for the first to execute. If you string 100 of these chips in series then you can execute something 100 times in the same frame!

The only issue is that the logic in any chip will only execute once per frame. So this limits you boundaries a litte. You wouldn't for example be able to create local variables that can be executed 100 times a frame because all techniques for local variables in Dreams, require looped signals or counter chips (so once per frame limitation).

You could use globals because they have nice interface gadgets. You can create duplicate interface gadgets at each duplication of your logic. But then you hit a wall again, Dreams only allows 128 globals!

The other option is to wire all your "locals" in series. So each local variable has a dedicated wire that feeds through each chip and looped into the next duplication of the logic. This could work. The internals of the chip can change that signal before passing it back out. The draw back here is hitting wire limits/thermo limits/port limits!

So in short, you might be able to execute emulation faster then Dreams update cycle but currently all methods have impacts either on thermo or on limitations of logic you can utilise.