# Was there an IBM computer where a memory upgrade was flicking a switch? **URL:** \ **Category:** Factual Questions\ **Created:** [May 30, 2024, 11:11am UTC](https://boards.straightdope.com/t/was-there-an-ibm-computer-where-a-memory-upgrade-was-flicking-a-switch/1002272 "2024-05-30T11:11:10Z")\ **Posts on this page:** 2\ **Page:** 2
**Author:** ![Tusculan](https://avatars.discourse-cdn.com/v4/letter/t/ecd19e/32.png) [@Tusculan](https://boards.straightdope.com/u/Tusculan)\ **Post date:** [June 1, 2024, 5:22am UTC](https://boards.straightdope.com/t/was-there-an-ibm-computer-where-a-memory-upgrade-was-flicking-a-switch/1002272/21 "2024-06-01T05:22:32Z")
Thanks everyone, this is just the kind of information I hoped for. it seems most likely that I (or the person who told me) misremembered and the issue was about processing speed. The additional information and discussion is also appreciated, interesting stuff. Incidentally, also good to see that ChatGPT was indeed making things up. ---
**Author:** ![md-2000](https://avatars.discourse-cdn.com/v4/letter/m/9d8465/32.png) [@md-2000](https://boards.straightdope.com/u/md-2000)\ **Post date:** [June 1, 2024, 12:16pm UTC](https://boards.straightdope.com/t/was-there-an-ibm-computer-where-a-memory-upgrade-was-flicking-a-switch/1002272/22 "2024-06-01T12:16:46Z")
> [@phs3](#): > > I never used an Amdahl (note correct spelling) but had heard that there was some version that was upgradable by cutting a wire. Still looking for a cite. IIRC our system engineer was the one who mentioned Amdahl had simply removed a board to provide the speed upgrade, and I found an article at the time that mentioned this - that there was a board to steal clock cycles to downgrade the computer. It made more sense than building two versions - after all, the components ran as fast as the components ran, it would be extra engineering to build something that ran _slower_ or buy two different speeds of components. I suppose they could have reduced the number of clock cycles, but I have no idea what implication that would have for peripherals. As I understood, the board would mimic an additional task running on the CPU to reduce processing power available for other tasks. > [@phs3](#): > > First solid-state memory was in the System/370 model 145 in 1970. When I started in 1980, we had a System/360 model 75 with a whopping 3MB of memory: one MB core, two solid-state. The big machine! Core was mostly made in Mexican _maquiladoras_. And some by Navajo. Always women, apparently—it was essentially weaving, and women had the skills for it. The articles I recall on core memory - describing how it was made, how it worked - mentioned it was made in the Far East (Thailand? Taiwan?). It was allegedly made by women because they had the threading skills and finer fingers, and in true patriarchal fashion, were probably paid far less. One of the systems analysts I worked with in the early 80’s knew assembler- he’d learned it because the 360 he’d learned/worked with COBOL on originally had only 40K of memory, and even with an optimizing compiler and overlays, some of his programs were too big. Memory cost money. I have the plans from _Radio Electronics_ for the first hobby computer (8008) around 1974 and it featured a memory board with 256 _bytes_ of memory, because that was the size of memory chips (multiple flip-flops) available. Like CPU’s, memory got bigger and faster every year through the 70’s and 80’s… > [@phs3](#): > > Make that “_always_ leased”, until the 80s. And yes, the competition between the divisions was brutal. Yes, it was a major industry and IBM owned it. When I first studied computers in the 70’s, the quote was “the whole world uses ASCII except IBM” and people laughed because 95% of the world’s computers were IBM. By the 1990’s, it was the same quote, and everyone laughed, because almost nobody used IBM mainfrmes or minis except as a legacy installation, and EBCDIC was such a clunky character code compared to ASCII. Indeed, the System/36 people were so upset at the power of the 12MHz 80286 that for a long time IBM only sold 10MHz PC-AT’s, thus ceding the PC market to Compaq who was not constrained by departmental competition and eventually marketed a 20MHz version. IBM considered PC’s a sideshow, so unimportant they let the development team use off-the-shelf parts, thus creating a clone-able PC and making the market wide open. They tried to put the genie in the bottle again with the PS/2 and proprietary expansion slots/cards, but it was too late. They tried to leverage their mainframe customers to push their proprietary tech, but the PC had spread so widely beyond that, it was a losing proposition. [Previous page](https://boards.straightdope.com/t/was-there-an-ibm-computer-where-a-memory-upgrade-was-flicking-a-switch/1002272.md?page=1)