Re: Interesting article in Spectrum about IBM's System/360
On 04/13/2019 09:11 AM, Jay Jaeger via cctalk wrote: For example, the IBM 7010 was an IBM 1410 done up in 7000 series technology (and was a compatible super-set of the 1410 and, via a toggle switch, the 1401). It had no architectural relationship with the 7090/7094, nor did the 7070 or 7080, near as I can tell. Yes, the 14xx were character-based decimal machines. The 7070 was a word-based decimal machine aimed at the business market. The 709x were word-based binary machines. Jon
Re: Interesting article in Spectrum about IBM's System/360
On 4/12/2019 1:15 PM, Eric Smith via cctalk wrote: > The article says: > > Poughkeepsie’s engineers were close to completing work on a set of four >> computers known as the 8000s that were compatible with the 7000s. > > > AFAICT, that is totally wrong. The 8000 series was completely INCOMPATIBLE > with any of the 7000 series machines. In fact, most of the 7000 series > machines weren't even compatible with each other, though the 7040 and 7044 > had partial compatibility with the 7090 and 7094. > > There are some 8000 documents on Bitsavers so you can see for yourself. > http://bitsavers.trailing-edge.com/pdf/ibm/8000/ > Furthermore, like the 8000 series would have been, the 7000 series (and the 700 series, and the 1400 series, for that matter) was more of a series of *technology* rather than a series of compatible computers. The 7000 series used SMS ECL (current mode), at least in a lot of places, whereas the 1400 series were essentially RTL with some DTL sprinkled in on the 1410. For example, the IBM 7010 was an IBM 1410 done up in 7000 series technology (and was a compatible super-set of the 1410 and, via a toggle switch, the 1401). It had no architectural relationship with the 7090/7094, nor did the 7070 or 7080, near as I can tell. >From "The Genesis of the Mainframe" by Bob O. Evans (an extract from a longer memoirs document, which was not itself published, to my knowledge) https://researcher.watson.ibm.com/researcher/files/us-bbfinkel/bob_o_evans_mainframe.pdf "Flush with the success of the 1401 and the 1410 in process — I was not willing to abandon those winners to join the 8000 series plan, which did not sit right with me in the first place because the 8103, 8104, 8108 and the 8112 were architecturally incompatible and I was certain compatibility was fundamentally important." "By May 1961 I concluded the 8000 series would be a serious blunder, in part because of the lack of compatibility within the systems family. I did not buy Dr. Brooks’ arguments that recompilation would be acceptable to make it possible for the programming from all the dissimilar architectures of existing products to work effectively on the dissimilar architectures of the 8000 series. There were other important reasons to scrap the 8000 series plan including technology choice. Jerrier Haddad backed my decision; the 8000 Series plan was killed." My experience with a couple of magazine authors during my career tells me that many of them do not understand much of what they are writing, and errors like this 7000/8000 thing are common. Another half truth in the article reads: "The power of compatibility was demonstrated in the fall of 1960, when IBM introduced the more powerful 1410 to replace the 1401. Software and peripheral equipment for the 1401 worked with the newer machine. " That was only true to the extent that the 1410 included a 1401 compatibility mode switch, which literally changed the logic so that it became a (somewhat faster) 1401. In its normal 1410 position, it could not run 1401 programs, and vice/versa. JRJ
Re: Interesting article in Spectrum about IBM's System/360
On 04/12/2019 04:14 PM, Chuck Guzis via cctalk wrote: On 4/12/19 11:15 AM, Eric Smith via cctalk wrote: The article says: Poughkeepsie’s engineers were close to completing work on a set of four computers known as the 8000s that were compatible with the 7000s. My tendency has been to consider 7000 xeries machines as transistorized 700 series. Certainly that applies in the case of the 7090. Well, to an extent. Yes, the 709x was able to run 709 programs, and had a few extensions. But, really, the hardware was VERY advanced. The 7094 was a real lightning fast machine, for the technology available at the time. In fact, it was faster than most of the 360 line that replaced it. But, the funny thing was, it didn't multitask well, and so you could only run one program at a time. And, spooling input and output to tapes slowed it badly (although not as badly as reading cards and printing directly would have). So, while fast, it didn't run efficiently. Slower 360's could keep busy by multiprocessing, and thus get more work done. Jon
Re: Interesting article in Spectrum about IBM's System/360
On 04/12/2019 12:41 PM, Carlos E Murillo-Sanchez via cctalk wrote: Building the System/360 Mainframe Nearly Destroyed IBM https://spectrum.ieee.org/tech-history/silicon-revolution/building-the-system360-mainframe-nearly-destroyed-ibm Yup, they bet the company on a new product. it was a VERY well thought-out bet, but still a big reach. One area they really made a mistake on was software. They designed a really ambitious OS (OS/360 MFT) and then an even more ambitious version (OS/360 MVT) on a poorly thought-out timeline. Fred Brooks actually had a nervous breakdown over it, and maybe some other guys, too. Fred Brooks' "The mythical man month" is just too short, and doesn't have enough actual anecdotes, but is a good read anyway. At the time he wrote it, there were probably a bunch of stories that he couldn't yet tell. Also, the hardware was a huge leap. IBM went from building computers with all purchased components on single-sided paper-phenolic PC boards to making their own transistors and diodes and packaging them on little ceramic hybrid modules, and then putting those on 4-layer PC boards. They pioneered a LOT of packaging technology on the 360. The developed flip-chip bump-bonding of semiconductors, and were doing this almost 20 years before anybody else were doing this. But, of course, there would be growing pains with such development. The entire state of New York was a bustling beehive of computer manufacturing. They made disk and tape drives, printers, hand-assembled close to 20,000 mainframe CPUs plus all the controllers and memory, between 1965 and 1969. Totally mind boggling! Jon
Re: Interesting article in Spectrum about IBM's System/360
On 4/12/19 11:15 AM, Eric Smith via cctalk wrote: > The article says: > > Poughkeepsie’s engineers were close to completing work on a set of four >> computers known as the 8000s that were compatible with the 7000s. > My tendency has been to consider 7000 xeries machines as transistorized 700 series. Certainly that applies in the case of the 7090. --Chuck
Re: Interesting article in Spectrum about IBM's System/360
Building the System/360 Mainframe Nearly Destroyed IBM https://spectrum.ieee.org/tech-history/silicon-revolution/building-the-system360-mainframe-nearly-destroyed-ibm It's an excerpt from a new book. I know the author. Very nice and smart guy who spent several decades at IBM.
Re: Interesting article in Spectrum about IBM's System/360
The article says: Poughkeepsie’s engineers were close to completing work on a set of four > computers known as the 8000s that were compatible with the 7000s. AFAICT, that is totally wrong. The 8000 series was completely INCOMPATIBLE with any of the 7000 series machines. In fact, most of the 7000 series machines weren't even compatible with each other, though the 7040 and 7044 had partial compatibility with the 7090 and 7094. There are some 8000 documents on Bitsavers so you can see for yourself. http://bitsavers.trailing-edge.com/pdf/ibm/8000/ Had the 8000 series in fact been compatible with the 7090/7094, I suspect that Bob Evans might not have recommended killing them, as part of the rationale for killing them was the fact that they weren't compatible with anything, but I think Bob still would have recommended that IBM develop a broad line of compatible computers (but mostly incompatible with 7094 and 8000) to replace them. I only met Bob once in 2004 at CHM, and only got to talk with him for a few minutes, so I could be entirely wrong.
Interesting article in Spectrum about IBM's System/360
Building the System/360 Mainframe Nearly Destroyed IBM https://spectrum.ieee.org/tech-history/silicon-revolution/building-the-system360-mainframe-nearly-destroyed-ibm