Re: [OT] Re: Continuous Reboots Plague Windows XP SP3 Users

Bill Ghrist wrote:
Tony Harding wrote:

<snip>

My first serious work with computers was in the early seventies as a field service engineer for Westinghouse Electric, working on power plant turbine generator controls and plant computers. We used at that time a computer built by Westinghouse itself--the P2000. What little programming I did initially was small diagnostic routines, which I wrote in machine language and entered bit-by-bit through sixteen backlit pushbuttons on the front panel (the application programs were written mostly in FORTRAN). The P2000 had up to 64 kword of core memory (16 bit), but no ROM of any sort. Software was loaded through a 60 cps punched tape reader, but the bootstrap program to run the tape reader had to be loaded manually through the front panel. We called the 60 cps reader "high speed" because some unlucky installations used a 110 cps ASR Teletype machine with tape reader for input. The next computer model--the W2500--actually had some ROM, which consisted of an array of diodes (diode=1, no diode=0) on a printed circuit card. These computers were controlling steam turbine generators (fossil and nuclear) putting out hundreds of megawatts of power. It may sound a bit frightening now, but they actually worked quite well.

The P2000 could use a drum memory, but few applications had them because of the cost. I don't remember how much storage the drum had, but I do remember that it was in a thick, sealed cylindrical case about two feet wide and a foot high, filled with helium. I got to open one up and replace some internal circuitry on a job in the boonies outside of Veracruz once.

Westinghouse also made an earlier model, the P50, which I did not work on, but which we continued to build well into the eighties. It was entirely discrete components (transistors, diodes, etc.). You entered the bootstrap program into these by touching a grounded test probe to printed circuit lands on the front of the PC cards. The P50 was used extensively for gas turbine generator control systems, and a few of them were still in use at the turn of the century--and yes they were Y2K compliant.

Interesting post, Bill, you were really "present at the creation" <g>. I get the impression, though. that you're confusing the drum (main) memory of the IBM650 with a drum attached & used like disk storage on other computers. I worked for IBM 1965-70 and never saw a drum (machine type 2301 for the S/360 series, IIRC) in the field - *way* too expensive for almost anyone (for those not familiar with drum storage, think of it as a hard drive with read/write heads for every cylinder, i.e., seek time = 0), maybe the military or NASA or some weapons research labs had them, no customers in my branch office did (and our customers included Rutgers).

Did a little Googling and found a page about a computing pioneer, Donald Knuth:

http://www.softpanorama.org/People/Knuth/knuth_biographic_notes.shtml

"The 650 could add or subtract in 1.63 milliseconds, multiply in 12.96 msec, and divide in 16.90 ms. The memory was a rotating magnetic drum with 2000 word (10 digits and sign) capacity and random access time of 2.496 ms. For an additional $1,500/month you could add magnetic core memory of 60 words with access time of .096ms. ..".

*60 words*!!! Why do we call the old days "good"? <vbg>

Interesting historical footnote, the 650 instruction format contained the address of the next instruction, which is not consistent with the von Neuman model, so it wasn't a true computer.

http://etl.luc.edu/gkt/luc/cs170/lectures/von_neumann_model

Von Neumann's model described instruction sequencing as most of us are probably familiar with, i.e., instructions are fetched & executed in a serial fashion (adjacent memory locations) until the sequencing is broken by a jump or branch instruction. The IBM 650 didn't process instructions in this manner at all.

<enough for now>
.

Quantcast