now one thing I've wondered is...how do processors
get to working? I mean
all it is is some metal with thin silicon inside...so how do some
transistors and diodes get it from some sheet metal to something that can do
all this? is it really just a billion little diode switch things which
literally act as binary switches (0s and 1s)? always wondered...thanks! :-)
The answer to that question could fill several books.
You are actually asking 3 questions at once, I think
1) How do you make electronic components (transistors, etc) on a silicon
chip?
2) How do you interconnect these components to make the standard, basic,
logic circuits -- AND gates, OR gates, flip-flops, etc?
3) How do you make a processor from said basic logic circuits?
The first thing to realiase is that there is nothing 'magic' about a
processor. It is just a (fairly) complex logic circuit. There;s nothing
much else in there other than gates and flip-flops, and it's possible to
understnad how a processor works at that level.
Most classic processors cna be divided into 2 sections (in fact in some
models of PDP11 the processor was 2 PCBs, and the split was into the 2
main sections).
The first is the 'Data path'. This contains the bits that the programmer
is normally aware of. The registers, the ALU, the swtiching circuits that
link the registers (and external memory to the ALU), etc.
The second, and IMHO, mroe intereting, part is the control section. This
is resposible for decoding the instructions read from memory, setting up
the data path to do the approropatie things, seqeuncing operations (like
reading a word from memory, incrementing the program counter, etc)
I've not found a good book on processor design at this level (there are
many terrible ones!) I learnt it the hard way be doing battle with the
schematics of a commercial processor and actually working out what each
gate did. TOok a few months of bedtime reading, but it was a great
feeling at the end.
-tony