elephant memory required? Been so long i'd have no clue what the specs were nor would i really give a crap.
i think it was an amd k5 or something
elephant memory required? Been so long i'd have no clue what the specs were nor would i really give a crap.
CP/M, S100 bus, 4,77MHz Z80, 8" floppy drives, P31 green phosphor CRT w/80 characters. A whopping 64KB RAM! That was about 1978 IIRC.
Way too much point to point wiring in that one too.
In the 80s I had a Heatkit H8 and HERO robot. Those were lots of fun.
In 1983 I acquired a Sanyo MBC550 microcomputer. It had an NEC V20 CPU and 256KB RAM. A good compiler and cheap compared to IBM PC!
I was a radar maintenance tech on Guam in 1950 -- just before the Korean war started -- and my CO was a Harvard graduate who received the Alumni news letter. In one of them was a paragraph or two about a machine that had been built -- or perhaps only designed at that point -- by two undergraduates, Kalin and Burkhart, which could evaluate the truth table for logical expressions. The example mentioned was checking the terms of an insurance policy for logical consistency. He showed it to me and asked if I knew what they were talking about. I didn't, but asked him to let me take it back to the barracks that evening. By morning I had designed relay circuits that could realize and, or, if then, if and only if, negation, exclusive or etc. The most it required were two DPDT relays for the most complex functions and a single SPDT for negation. He asked if I could build such a machine and I said yes -- given enough relays. Aircraft used 28v relays so he got on the teletype and requisitioned from every supply site from Honolulu to Tokyo their stock of 28v DPDT relays. They must have thought that every aircraft at Anderson AFB had been struck by lightening to kill so many relays. Using two ganged telephone rotary stepping switches I wired up a 10 variable input -- i.e. it had ten output wires that would sequentially step through the 1024 states for 10 logical variables. Using the 28v aircraft relays I constructed modules for a large number of the functions that could be wired using pin jacks and pins from the front of the console. You set in the logical expression whose truth table you desired to map, turned on the stepping unit at state 0 --- 0 and let it step it's way through the 1024 states. When a state was reached for which the logical expression was true, a current flowed through the circuit closing a relay and stopped the stepping switches. You could then copy down the values for the ten logical variables from the state of ten lamps on the front. Pushing a button then caused the stepping switches to resume. At the end, you had the truth table for the original logical expression. It was the most satisfying computer I have ever built -- with all those relays clicking in and out and lights flashing as states changed. We named him George -- and when I returned from the South Pacific in 1953 managed to ship him home as hold baggage and talk my way through the port inspection. The secret was that George didn't look like anything they had ever seen and they were mostly looking for people trying to steal governenment property. He entertained a whole generation of student engineers just as digital computers were coming on the scene.
I honestly don't know how you can remember specific details of your most ancient machines.
all of you people saying "way back in 2004 i built some athlon 64" make me sad because i still use mine as my main. it's really not that bad! text