A super minor nitpick: it’s jarring to see the Amiga referred to as 16 bit. It wasn’t described that way at the time: it was universally (that I saw anyway) called a 32 bit machine, and reasonably. It had a flat 32 bit address space (although the 68000 itself didn’t support all those address lines because what kind of supercomputer would need 4GB of RAM?). All the registers and operations were 32 bit. Some of the internal operations were implemented in 16 bits, but that was invisible to programmers. Newer models with definitively 32 bit CPUs like the 68060 were nearly 100% backward compatible with older models at the CPU instruction level, even if newer OSes weren’t backward compatible at the API level. In fact, the only program not forward compatible at the instruction level that I remember offhand was Microsoft’s AmigaBASIC. It used the top bits of pointers to store data because the 68000 would ignore them when accessing RAM due to that lack of address lines.
I just don’t see a way to justifiably call the Amiga a 16 bit machine. Although the A1000 had some 16 bit hardware paths, a maxed out A3000 definitely wasn’t 16 bit, and they were nearly completely compatible with each other minus newer features.
Amiga was full-on 32 bit machine. It’s weird to hear it called anything else.
I just don’t see a way to justifiably call the Amiga a 16 bit machine. Although the A1000 had some 16 bit hardware paths, a maxed out A3000 definitely wasn’t 16 bit, and they were nearly completely compatible with each other minus newer features.
Amiga was full-on 32 bit machine. It’s weird to hear it called anything else.