Re: i5 SFF Desktop / CUDA


First program I wrote was on an IBM 7090 at the University of Michigan using the "MAD" language. Seriously, MAD, Michigan Algorithm Decoder, was the language. It was all punch cards in 1963. I decided computers were fancy wastes of time that had potential for the future; but, I had radios to design in my future so I stuck with learning a solid EE background specializing in courses around communications uses of EE. Along about 1970 I found myself at Rockwell International in Anaheim and in possession of a genuine first generation HP-35. I still have it with the first generation "ln 2.02" error still in place. I could not spare the tool long enough to get the free fix. I had my fingers programmed for circuit analysis on it already. About a year and a half later I found myself in a 6 "man" office with me, three men, and an HP 9825 programmable calculator with a plotter - sitting idle -- too long. I was sitting idle for short periods. I mumbled to myself, "I wonder if I can do some simple circuit analysis on it."

So over a few months I hacked together a simple minded "ladder" sort of circuit analysis program. (Assign a load resistor and a voltage across it. Connect the next component and figure out what it's input had to be to create that output. Go backwards through the circuit to the input. Renormalize the results and plot it.) At about that time we were building the group's third Hedy Lamar radio. I was building the RF/IF systems. We had a 13 MHz wide 70 MHz first IF. It had a voltage variable capacitor tuned IF. The caps were inserted back to back in series to minimize capacitance variation with voltage. The twin tuned circuits were inductively coupled - at the centertaps of the capacitors. I missed a spurious circuit path in my initial cut at the design. The ladder analysis found it. I de-Q-ed the stray path with a small resistor until it was satisfactorily far down. A very talented tech built the circuit. When tested the results were within a dB of what the plotter gave me. Can you spell convert? "Hey, this digital thingus dingus has something to say for itself."

The program ended up migrating to an HP 2011S computer with DITEK (sp?) tape decks. (messy electrostatic) Versatek printers, and a software suite that included a linkable BASIC interpreter, a FORTRAN compiler, and an assembler. I ended up with source code for the OS, drivers, and BASIC interpreter. I added circuit analysis as a command in BASIC. (And I also ended up scaring an HP tech who had dismissed me as merely a guurrlll. The computer had a fault. I told him where it was. When he did it his way and got to the problem he looked at me sorta funny. He was man enough to believe me in the future - good guy. I liked him.)

The first one I owned, ignoring the HP-35 of course, was a Processor Technology SOL-PC. By the time I had it fully tricked out PT would not have recognized it at first glance. Would you believe an S-100 bus computer in a 3" high box about 12" wide and 18" deep - with three cards plugged into it?

Those were fun days - I was designing RF hardware and some limited software all at the same time putting some of it into GPS Phase IIB satellites. It was downhill on the dark side (digital) from there. Some others have me on age and length of time playing seriously with computers - but not many. (Loren, my partner, is the champ as I see it. He was working at Burroughs by the 1972. Within a year he was working on their MCPs, Master Control Programs. He's still at it working on the UniSys A series emulators. He's a hairy chested big machine guy. I consider myself a dabbler in comparison.)

{^_-} There is one good thing about being a white hair - you get to brag.

On 20190704 04:18:53, Simon Brown wrote:
You win – your callsign sort of confirms you would win anyway (I’m 1957 vintage myself). My first computer at university used punched cards, then I progressed to paper terminals, then my first monitor!
How it’s changed…
Simon Brown, G4ELI <>
*From:* <> *On Behalf Of *Allan Isaacs
*Sent:* 04 July 2019 12:03
*Subject:* Re: [SDR-Radio] i5 SFF Desktop / CUDA
I was a Plessey programmer on an XL9 system at GCHQ back in 1970 Simon.
If I recall correctly it was mini-Coral and Exec 9.. no USB sticks then.. just paper tape
We made the XL9 at Exchange works. It was developed from the XL4 used to supply the 4 minute warning.
That system had 1000 seven foot racks and I think used no more than 20 computers.
All germanium transistors …..
Allan G3PIY
*From:* <> [] *On Behalf Of *Simon Brown
*Sent:* 04 July 2019 11:57
*To:* <>
*Subject:* Re: [SDR-Radio] i5 SFF Desktop / CUDA
Sounds just like my 1996 first Windows PC which I still have – it’s up in the attic! It cost so much I can’t bring myself to throw it out, even though it’s not been used for at least 20 years!
And the first computer when I started coding for myself rather than being a wage slave is still in use – a VAX station 3100!
All that money, where did it go?
Simon Brown, G4ELI <>

Join to automatically receive all group messages.