ok, so things are getting better right now. technically, a lot better.
so, today i went to bed around 5 am, and then i woke up at 9 for a meeting with my comparch prof. basically, she asked us "do you think you'll be able to finish the project", and we were like "we should...but no guarantees". and yeah...
i worked on that damn thing for over 12 hours today. but it is finished. it works...and gives results that are true to the mark. just in case you're wondering what i'm talking about, i made a processor. according to code, it runs just under 14 MHz, but that's not too technical for a speed...because the program we use to clock it is really inaccurate. I think it could probably run in over 50 MHz, but i doubt i'll try that anytime soon. i've heard people say that their projects in the past have ran in the 100+MHz zones, but screw that. they cheated...we actually built the whole thing from the ground up and used only two pre-synthesized parts (the memory and the register file), but even those we edited by hand to produce what we needed out of them. in past times, people have used pre-synthesized product packages that basically make the architecture for you...which is not that hard. i personally built the ALU (arithmetic logic unit), the barrel shifter (shifts bits around based on an input), a whole shitload of multiplexers (they choose which input to output based on a control signal), a register file (holds information for multiple units over a long period of time) and a bunch of register components (temporary storage units). i may have made some other things, but they were mostly easy. altogether, there are exactly 6911 logical gates used within the processor...but about 500 of those are redundant (because the program is retarded and you can't have a wire with two names to connect two different connections). i probably put in about 40 hours in just making those units...not to mention the assembler i made (which is almost complete...but it won't get turned in because it's in Java, and Java is the devil). then i also compiled the thing (integrated all of the parts from separate units into one big thing!), which took well over 40 hours this week. it's been a long week...and i'm glad to see some rewards.
when the processor first worked outside of 5 or 6 clock cycles, kober and myself rejoiced for the better part of a minute...until we realized that it crashed again a couple of cycles later. there were a lot of small errors, but some other ones were changed to make things work...because without the change they wouldn't work at all. boy am i glad this is over!!! we have to present on Monday...
...but as for now i'm off to bed. i only got 3 hours basically between yesterday and today, and i hope to at least double that before i have to turn in the Disco take-home test that we got on monday. man...school is forever busy. but boy do i like the challenges. plus, did i mention that i frieking built a processor!!!