Hi That's a bit past the register sized range, but still a bit small by today's standards.
Bob On Jun 23, 2013, at 5:50 PM, Paul Berger <[email protected]> wrote: > Hi: > > The SAGE computers, which I had the pleasure of seeing the last two > operating, had an all vacuum tube array of core that consisted of 33 planes > of 64 x64 cores for about 16K worth of memory. These where all vacuum tube > computers. IBM offered a 4K all vacuum tube core storage unit for the 701 > and 702, the same unit was built into the 704 and the 705 had a larger core > storage with 35 planes of 50 x 80 cores. The Remington Rand Corp. and the > RAND Corp. also shipped computers that used core for main storage in the mid > 50s which likely used vacuum tube drivers. At that time there where > apparently no transistors available that could supply the drive current > required for core memory. > > On 6/23/13 5:29 PM, Bob Camp wrote: >> Hi >> >> I've both used and worked on core memory machines. They ones I have seen all >> used solid state devices in the core memory sections of the machine. I've >> never heard of a pure tube machine with more than "register sized" core. >> >> Bob >> >> On Jun 23, 2013, at 1:47 PM, Chris Albertson <[email protected]> >> wrote: >> >>> Magnetic cores were not invented until the 1950's and realy cam into use as >>> tubes were beibg replaced by SS. But there isnot reason yu can't build a >>> tube computer with core memory. I have actually seen and used a computer >>> that had one megabyte of core memory. The stuff was still in use in the >>> late 1970s 1MB was a lot of RAM in 1975. >>> >>> You can have very good reliability with tube circuits. It was just that >>> few people wanted to pay for it. Down time was cheaper. It is not hard to >>> add redundancy to a circuit but it does have a huge cost multiplier effect. >>> 4x or 5x the price. One simple way is to use 3 or 4 tubes with their >>> output tied to a resistive adder. If one tube fails the result (because it >>> is binary) is still the same. With computers no one would pay for fault >>> tolerant design until it was reasonably affordable. Even today we mainly >>> just put up with failure except for airplane controllers, huge web sites >>> like Amazon and the like. >>> >>> >>> On Sun, Jun 23, 2013 at 9:53 AM, Brian Alsop <[email protected]> wrote: >>> >>>> On 6/23/2013 14:40, Bob Camp wrote: >>>> >>>>> Hi >>>>> >>>>> AC137 doesn't ring any bells. True tube core (no solid state at all) >>>>> isn't something that was dimensioned in K words. A couple hundred words >>>>> was >>>>> pretty big stuff. "Quite a bit" of core done that way is a lot of tubes. >>>>> As >>>>> the number of tubes goes up, the time to failure comes down….. hours … >>>>> minutes … who knows. >>>>> >>>>> Bob >>>>> >>>>> >>>> Yeah, it gets to be like the cross country aircraft races in the 20's. The >>>> mechanic had to fly with the pilot. (The MTBF of many of the engines used >>>> was measured in hours.) If necessary he had to climb out on the cowling >>>> while in flight to change plugs and fix whatever possible without landing. >>>> What would OSHA say about that? >>>> >>>> Needless to say future generations will probably find lots of aircraft >>>> spark plug artifacts in their digs. >>>> >>>> Brian/K3KO >>>> >>>> >>>> >>>> ----- >>>> No virus found in this message. >>>> Checked by AVG - www.avg.com >>>> Version: 2012.0.2242 / Virus Database: 3199/5932 - Release Date: 06/22/13 >>>> >>>> >>>> ______________________________**_________________ >>>> time-nuts mailing list -- [email protected] >>>> To unsubscribe, go to https://www.febo.com/cgi-bin/** >>>> mailman/listinfo/time-nuts<https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts> >>>> and follow the instructions there. >>>> >>> >>> >>> -- >>> >>> Chris Albertson >>> Redondo Beach, California >>> _______________________________________________ >>> time-nuts mailing list -- [email protected] >>> To unsubscribe, go to >>> https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts >>> and follow the instructions there. >> _______________________________________________ >> time-nuts mailing list -- [email protected] >> To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts >> and follow the instructions there. > > _______________________________________________ > time-nuts mailing list -- [email protected] > To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts > and follow the instructions there. _______________________________________________ time-nuts mailing list -- [email protected] To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
