>It helped solidify many of the concepts from nand2tetris. For some reason doing it all physically with real chips and wires made it all a bit more memorable.
Hang on a second, does nand2tetris not involve real chips and wires?
This is largely about accessibility. If it’s tied to hardware, fewer people can do it. Additionally, real hardware means there’s additional opportunity for error due to faults in the hardware or environmental variables.
That said, Nand2Tetris could, in theory, be done with a bunch of 74xx NAND chips on a breadboard.
The Nand2Tetris computer is a bad fit for 74xx chips and breadboards. It's 16 bit, for one thing, which implies a lot of chips and a lot of wires on a lot of breadboards, and the usual current problems that come with that. Better have a good source and an oscilloscope nearby. Also, the virtual design sweeps a lot of important timing issues under the rug, so you need some EE chops to translate it correctly to real hardware.
I know of a few people who managed to get a working Nand2Tetris computer on breadboards, but it took a lot more time and money than they thought it would.
Hang on a second, does nand2tetris not involve real chips and wires?