Interesting that processors are getting wider and wider, whilst (perhaps not in the same timeframe) we have moved away from parallel interfaces towards serial ones. I know there are reasons for that in operations-per-cycle and the difficulty of synchronising wide busses off-chip but I wonder if those sweetspots will change again.
On Fri, Sep 1, 2023 at 6:53 PM Rick Bensene via cctalk < cctalk@classiccmp.org> wrote: > Just to add, interestingly, Singer also purchased General Precision from > Librascope. > Librascope/General Precision were the folks that had earlier acquired > Royal-McBee. Royal-McBee developed the wonderful (some consider the first > "personal" computer) LGP-30 vacuum-tube, magnetic drum computer that was > designed by Manhattan Project theoretical physicist Stanley Frankel. > > Frankel had quite a legacy in the world of computing, having contributed > to the design of the delay-line-based Packard Bell PB-250(with Max > Palevsky), and development of a custom high-speed computer for Continental > Oil Company called CONAC (used for data reduction of sounding operations > search for oil deposits). > > Frankel also developed an early electronic calculator design that was > purchased by Smith Corona/Marchant (SCM) and produced as the CRT-display > SCM Cogito 240 calculator, augmented with Square Root as to Cogito 240SR. > > Frankel also collaborated with SCM on the development of the logic for the > first set of LSI integrated circuits that were used in the later Nixie-tube > display Cogito calculators. > > He also developed a very interesting calculator, based somewhat on the > principles of the LGP-30 computer for Diehl in West Germany. The machine > was fully transistorized and used only 142 transistors in its logic. It > was based on magnetostrictive delay lines (two of them), and was a fully > microcoded architecture, I believe the first electronic calculator to be > completely microcoded. > > Since read-only memory (for the microcode) was either physically very > large, or complex and expensive to build at the time (diode ROM, wire rope > ROM), the microcode was loaded into the calculator at power-up time from a > two channel punched metal tape. One channel provided the clocking, and > the other channel provided the bits. > > It took just under a minute from when the calculator was powered on until > the microcode was loaded into a delay line, and from there, all operations > of the machine were controlled by the microcode in the delay line. > > The machine was able to be implemented with so few transistors because the > microcode word was quite wide, and was designed so that it was sequentially > interpreted as the bits streamed out of the delay line, so not all that > many flip flops were needed. Working registers were stored in the other > delay line, along with program steps (yes, the machine was programmable). > > The design was very elegant. The machine debuted as the Diehl > Combitron, and the cool thing about its design was that it was really easy > to augment by just changing the microcode tape (which was quite easily > done...bugfixes could be easly installed even by end-users, though such was > discouraged). > > Soon after the Combitron was introduced, an augmented version was > introduced called the Combitron-S that added a small amount of I/O > circuitry and additional microcode to implement operations to allow the > addition of an external punched paper tape reader/punch. > > An interesting aspect of electronic calculator history is that there are a > number of people whose names pop up at various points in time during the > evolution of the technology. Frankel was one of those, along with a cast > of a few others, all of whom had major impacts in the realm of electronic > calculator (and the eventual evolution of the electronic calculator into > what became the microcontroller/microprocessor that spurred the development > of the personal computer). > >
