• Please review our updated Terms and Rules here

Thoughts on minicomputer to microcomputer transition (PDP to Macintosh)

Hardly. Single-board computers were around before the Apple-1. Consider, for example, the MMD-1. There were also development boards, such as the Intel SDK boards. Note that the KIM-1 has an expansion bus.
 
voidstar78 said:
Wasn’t the innovative factor of both the kim-1 and apple-1 was that they were a single board? And to whatever that meant exactly, didn’t that ultimately drive down cost?
Click to expand...

That's the lame as all heck excuse some highly motivated jobs fanboi with an IEEE membership card tried to throw out there to try to make the case the Apple I was somehow special, but realistically it was a flop in terms of both technology and price. The Apple I was just as expensive as better computers on (*GASP!*) two boards and they sold a whopping 200 of them, so people really need to get this fake story about it being some kind of price/innovation leader firmly out of their heads. That it was important in any way other than being the first product of Apple Computer Inc. is fan headcannon, not history.

Also, if we're really going to nitpick here: A: the Apple-1 need a second board plugged into it to save programs or do any communication with anything but the TV (which its brain-damaged architecture can only do at a double-digit number of characters per second), and B: it came as a motherboard which the buyer had to add all the other parts to, including power supply transformers and a bare keyboard that might need hacking to make work. So how is it not also a kit? Granted since *most* of the soldering is done it's more of an Ikea headscratcher than a 1000 part Revell aircraft carrier, but it's still not even close to something like a PET or TRS-80.

Final edit, I swear: (rantrantrant) It also didn't have a working BASIC at launch, and when it finally did get BASIC it was never finished. It doesn't even have program save/load commands, the only way to save a BASIC program is drop to the ugly machine language monitor and do a manual memory dump. Using a TV is nicer than a hex keypad, sure, but out of the box it's capable of nothing the KIM-1 can't do for a third of the price. Hook a terminal to a KIM and it even has a better machine language monitor. Again, people who claim it's the first "personal computer" over the much better machines you could buy at the time for not much more money only because it came *mostly* on one sheet of fiberglass have no idea what they're talking about.
 
Last edited:
Oh it’s clear both apple-1 and Kim-1 were kits, not ready out of the box systems of any sort. But they were the starting points to what did become successful products a year (or two) later. All of these systems had a rough looking prototype a year or two before the product (eg the odyssey had the brown box, the coco had the green thumb). It’s like those early cars with no windshield and bicycle wheels. Some of those prototypes are better archived than others.

Would you agree that the KIM-1 did show that a 25$ cpu could enable a viable system ? ( as opposed to the 200$+ alternatives at the time ). And I can't find numbers on how many KIM-1's were sold. I agree due to being a "grossly incomplete kit" and low "sales" (it wasn't quite a sincere product to begin with), we can nix the Apple-1... But the KIM-1 was somewhat complete (not as a personal computer -- although it did somewhat have that potential, but I mean more for some industrial control purposes), and perhaps over a 1000 were sold?
 
Last edited:
krebizfan said:
The election of 1972 was conducted on punch cards in the LA region.
Click to expand...

Do you know happen to recall if that was a "new" system? In any case, on the merit of price/performance, looks like punch cards were still a 70's thing (although I'm not aware of any micro that used punch cards as an input?? in a roundabout way, maybe the 5100 did when started up in terminal mode to an IBM system -- wouldn't that be a way to adapt punch card data over to a tape?)
 
It would be difficult to categorize the KIM-1 as a "personal computer"--perhaps as a "6502 evaluation board". The HP65 calculator (1974) was portable, had mag card I/O, full keyboard and was the first programmable to make it into space.

Please be honest--are you picking our brains for a school project? How many years do you intend to devote to this thing (i.e. do real research)? Inaccuracies in tech histories is one of the reasons that I canceled my subscription to IEEE Annals.

I recall running across an old post in some technology forum where a guy from Europe said that he was completely unable to find any information at all on a massive mainframe system. I wrote him and offered, as one of the developers, that I could tell him anything he needed to know. His response was "Oh, that was 4 years ago and I've moved on." Clearly he was an academic writing papers and really only wanted to know what might give him a passing grade.
 
voidstar78 said:
Do you know happen to recall if that was a "new" system? In any case, on the merit of price/performance, looks like punch cards were still a 70's thing (although I'm not aware of any micro that used punch cards as an input?? in a roundabout way, maybe the 5100 did when started up in terminal mode to an IBM system -- wouldn't that be a way to adapt punch card data over to a tape?)
Click to expand...
Fairly new. The Votomatic was introduced in LA County in 1968 but continued to spread around the nation reaching a peak in the early 90s. I remember it because the voting officials had a practiced routine of having a grade school child produce a sample ballot in order to teach the adults how to use the system without the adults being embarrassed.

Almost all micros could use punch cards for input; the punch card reader was a fairly standardized serial device. The problem for punch cards was the size and cost of the punch. Sometime in the late 70s, the CRT terminal became cheaper than the desktop card punch typewriter and as the punches failed, the switchover happened. I think card readers remained with the voting systems and educational systems where manual punching happened. Some punch card readers could also read mark sense cards which had long run.

DEC had a special format mark sense card that was used for programming. Much cheaper than having enough proper punches to handle an entire class. See https://www.pdp8online.com/shows/vcfe14/vcfe14.shtml?med for an example. Read the instructions picture and imagine how difficult that would be during the high pressure of a test.

Paper tape had a larger use with micros since booting routines to load from paper tape were common. Of course, most of the punch-readers had been withdrawn from service along with tele-types so the used reader was very cheap. Paper tape also survived for a time since that was the cheapest way to get PDP software from DECUS.
 
Nope, no project or paper. Just passing time during storms and recession. Too cold to go jetski. As posted originally, I'm not so much interested in identifying any "first" - but just musing over the overall flow of the transition (at least the highlights, or insofar as to what can fit in a typical page or screen), as there was a rather deliberate vision to "a computer in every home" (followed now by a computer in every pocket). My daughter asks about the old systems now and then (we fire up old Wheel of Fortune on the holidays), but I doubt she'll ever do a project about them.

We can rent time on Quantum computers now - I was trying that out over the summer - which just made me think it's like we've come back around in a circle again: we used to rent time on mainframes. Decades from now, maybe we'll be developing PQC's (personal quantum computers). But we're just starting to explore how to program those things, and we'll see what useful applications emerge.
 
Last edited:
Great info about the Sphere1. I wasn't sure if that was for real, but they had a "factory", user group newsletters, CTRL-ALT-DEL legacy (and reading about the 6800 led me over to the Tektronix 4050, cool stuff). I hope a functional Sphere is found someday.

Also neat about the HP-65 (with the little program card inserts).

Thanks for pointing those out!
 
voidstar78 said:
Great info about the Sphere1. I wasn't sure if that was for real, but they had a "factory", user group newsletters, CTRL-ALT-DEL legacy (and reading about the 6800 led me over to the Tektronix 4050, cool stuff). I hope a functional Sphere is found someday.
Click to expand...

The Homebrew Computer Club newsletter for September 1975 relates that the reps from Sphere who were in town for the Wescon 75 trade show (where the 6502 was introduced) dropped by a club meeting and gave them a demo of both the Sphere and of some kind of prototype color graphics device. So yes, it was real.

(And I guess Steve Wozniak must have missed that meeting if he actually believes that flaming BS he wrote in iWoz about how revolutionary-first-time-in-history-blaw-blah it was when he flicked the power switch of his computer… which to be absolutely crystal clear, was at least a month after this demo of a basically finished commercial product.)
 
Okay, I'll offer this then. Get off the Apple thing if you're trying to take an unbiased view of history. Up until at least 1985, the place where you were most likely to find an Apple computer was in a classroom. Why? Because Apple offered sweetheart deals to schools, particularly in a one-size-fits-all package. When Apple tried to upgrade its platform to meet the PC with the Apple III, it was a dismal failure. In the early 90s, I witnessed schools dumping pallet-loads of Apple II systems for recycling. You could probably have a ton or so by simply offering to haul them away.

Even at that time, the PC-and-compatibles market held most of the personal computer market. Up until the "invasion of the clones", IBM held probably 70 percent of the personal computer market.

The early Macintosh 68K systems were hampered by their small display, but aided by the fact that you could pack one into your luggage. Using technology from Xerox PARC, they made the GUI their standard, which was a great move.

But Apple staunchly resisted the demand for laptops for a very long time. Indeed there was at least one company who would take a Mac and repackage it into a portable version. Apple probably would have gone bankrupt during that time had it not been for Microsoft lending some money.

Apple was perhaps unique in suing anyone trying to clone their hardware and software and making sure that the ROMs didn't wind up on the surplus market. I recall a bin at Halted with Mac 128K boards for $5 each--and no ROMs. In other words, Apple achieved its success by stifling competition--an attitude that prevails to the current day.

My point is that Apple's rise had little to do initially with technical excellence, but rather with anti-competitive practices and sales strategies.

At least that's how I saw it--and I was there.
 
I was in those classrooms full of TRS-80s, Apples, and Commodore PETs (depending on the region). That's why I have a PET today, is because I found it in a dumpster. And the reason they ended up in dumpsters is because they were being replaced by IBM PCs (as hand-downs from the local university - those IBMs had big stamps on the side marked from the CIS departments they came from). And I fully remember those donated systems being tax write-offs (to Apple, etc).

One of my neighbors was founder of Personalized Programming, starting The Medical Manager software on a $600 TRS-80 Model 3 (that eventually becomes WebMD decades later). And we all know the Sierra OnLine story, that started from the Apple 2 (pre-1980), the "killer app" of VisiCalc that finally made the Apples' appealing (also pre-1980). So those systems are all relevant to the story (maybe not in any technical innovation, but in having "market appeal" and helping to increase awareness that a home-computer could be used for way-more than single-screen gaming). Even if an earlier system was technically capable of it, it needed to achieve a "critical mass" of users (to make further development in software worthwhile).



But you're right, Apple "barely' made it... We don't have exact figures, but I think these ballpark estimates are reasonable:

Code: 
System         Units Sold
PDP-11          600k
PDP-8            50k
Wang 2200        73k (A/B only)
IBM 5100         50k (not inc. 5110 or 5120)
Altair 8800      25k  (kit and prebuilds?)
SOL-20           12k
TRS-80          200k  (FCC issue stops prod)
Commodore PET   200k  (by 1980)
Apple 2          64k (1977)
Apple 2+        500k (1979)
TRS-80 M3       100k (1980)
IBM PC 5150    3200k (1981)
IBM XT         2100k (1983)
Apple 2e       4200k (1983)
IBM AT          900k (1984)

By these numbers (which we have to take with a grain of salt, since it's unsure if they represent all years of production or not for all cases), out of the gate the Apple 2 struggled - since it's under half of its two contemporary rivals. But once you get over 100k, that's a pretty meaningful user base.


The original Apple reference manual seemed fairly open about the system.

Nintendo was also fairly aggressive in its business practices as well, perhaps more so than Apple.




I didn't realize Apple was so "late" on making a laptop (I never owned any Apple computers), I was going to try adding one but it doesn't fit my 1984 cut off.


1664410578009.png

Also, Compaq did make an early 286 clone, but not until 1985.
I squeezed in a Sun-1 -- which I think were something like $8000+, so not exactly the humble home devices (hence "workstation").
 
Last edited:
I was reading through the Radio Electronics 1974 article about the Mark-8 "minicomputer", which has a part list at the end. That stack of boards is akin to what is in the IBM 5100 (but the 5100 has them attached to a "mainboard" that acts as a kind of bus across all the boards -- and I'd assume that general arrangement used in the 5100 was also in their 1973 prototype, but I've never seen photos of the SCAMP opened up).

1664445558726.png


I don't care to nitpick on who "developed" the single board first -- because to be fair, you'd have to look at viable prototypes -- also, it's an organic concept, as one arbitrarily decides what to include on that single board (along with the process they choose to come up with the arrangement and technique of applying the components). Multiple solutions were evolved across regions, so I could see how locally (pre-Internet days) they could each believe they were the "first" (to their set of choices on what to include, and to document it to make it repeatable and understood). So each of those early boards deserves some kudos on recognizing the need for a single board, and then doing it.

Anyhow, I thought it's also fun to see how the "minicomputer" stack of 1974 looks similar to the modern day stack of processors used in quadcopters (or at least those little racing ones).

EDIT: Interesting how "mark-8" is still using the term minicomputer. Would the 5100 also still be considered a minicomputer?? I can't recall if the Wang had memory mapped video, but I know the 5100 does (1K @ 0x0200). But just see the images - you can see the difference in "bulk" and "complexity" at "whatever was before single board" vs what became single board design. (I think the Mark 8 was "double sided" boards, the 5100 just has components on one side)
 
Last edited:
voidstar78 said:
So each of those early boards deserves some kudos on recognizing the need for a single board, and then doing it.
Click to expand...

No, they don’t. The reason a “full” computer from 1973 is inevitably a thick PCB sandwich with extra mustard while it was possible in 1976 to squeeze all the moving parts onto one open-face board is entirely thanks to innovations in semiconductor manufacturing. Nobody at Apple or anywhere else gets a “I’m a smart boy” warm fuzzy for “recognizing” the idea of using the resources they have available.

…. Actually, since MOS actually made the chips for the KIM-1 maybe they get extra credit. The RRIOT chips it used were kind of the forerunners to the VLSI ASICs that took over computer manufacturing in the mid-80’s.
 
I was impressed with the amount of functionality the MCM/70 fit into a portable package. But integration ramped up quickly. By 1976 it was possible to make a working system from an 8085 and a support chip.
 
Also, FWIW, look up a photo of, say, the bare Sphere 1 boards, like the ones in this article, and compare them to the Apple I. Yes, it's two boards (plus a keyboard) instead of one (plus a keyboard), but the boards are smaller. The packing density of the two systems is very similar.

(The reason S-100 systems like the Altair are so relatively huge is mostly a technical oddity that has little to do with their level of digital integration, it's because they were designed to require every circuit card to do its own power regulation... and also the bus is a little weird in a way that would seem pretty outmoded within a couple years but kinda made sense in 1974. Shrinking things like power supplies was reliant on advances in devices like switching PSUs, which were kind of new at the dawn of the 1970's.

But even within those parameters you could make a pretty small computer out of S100 cards. SOL-20 or Poly-88 are good examples.)

And here's the cool thing about the Sphere: late model Spheres came with an 80x24 video board instead of the 32x16 board of the original. Which was something they could do because the system was modular, not crammed all on one board. Isn't that, um, exactly one of the big reasons the IBM PC architecture ended up crushing the single-board competition by the end of the 1980's, it could evolve with changing requirements? Packing everything onto a single PCB board is NOT intrinsic to what defines a "Personal Computer". Being small and cheap, yes, but small and cheap can be solved multiple ways with different tradeoffs.
 
I tend to put the machine that made it obvious that much smaller boards would work was the HP-35 which was part of a plan to place the entire HP-9100 in a pocket sized machine. Okay, it wasn't until the HP-67 that the pocket calculator had both the magnetic card and more than 200 program steps matching the HP-9100A. The microprocessor wasn't going to stay a slow inferior cute toy in comparison to TTL for much longer. The mid-70s trainers were effectively oversized calculators built around general purpose microprocessors replacing more complex demonstration designs using the same microprocessor.

https://www.hpmuseum.org/tech9100.htm The 9100 It has a picture of the 5 boards that plug into the lower board with about 600 components.

https://www.keesvandersanden.nl/calculators/hp35_inside.php HP-35
https://www.keesvandersanden.nl/calculators/hp67_inside.php HP-67
 
I tend to say "single board computer" in quotes because it's inconsistent on what that means - since exactly right, it means a specific choice on what gets integrated together, and that choice impacts the cost, how long the system remains useful (how quickly it becomes obsolete), and the form of the cabinet (which becomes a kind of "personality" of the system). Some choices didn't pan out (TI's attempts, maybe?), some were wild success (market-wise at least). Note, Kildall would say the Apple2 was saved only by disabling all the 6502 component and using a Z80 SoftCard. Luckily it had the expansion capability to do that :) (maybe, maybe not - but I think more to the point was: software quickly helped shape what systems became a success - non-technical users wanted to jump to their application, not just boot up to BASIC)


Every SOL-20 I've seen is packed full of cards. For whatever reason, the SOL-20 didn't become the computer in every home - reliability issues? lack of internal storage? I'm sure lots of other reasons could be proposed. But I'll agree to your point, neither did any of the Trinity systems - but they were the next step, at widening the awareness of what these systems could be used for (by reaching below $1000). To me, the Trinity collectively represented the Model-T moment - Ford wasn't making the best car, but it was affordable and appealing.



The MDM/70 is neat, but with that single row screen, it seemed the HP advanced-programmable-calculators (of about the same years) were about the same (but without APL). Interesting that they switched to AMDs (which I guess makes sense, AMD were also a Canadian company?), but they kept going with those odd screen sizes. I think I saw the MCM/70 also had a modem port?
1664504275882.png
EDIT: Don't think anyone meant to imply the Trinity originated the concept of a single board system. It's just that they picked a particular combination of parts (yes "off the shelf existing" parts), and gambled on whether that translated into an appealing system. They each got a couple football stadium worth of customers, so that's a start.
 
Last edited:
Interesting that they switched to AMDs (which I guess makes sense, AMD were also a Canadian company?)
Click to expand...
Do your research. AMD was the brainchild of Jerry Sanders and his cabal of Fairchild engineers in 1969. Initially, Santa Clara, then Sunnyvale, California. (Bob Noyce and Gordon Moore were also Fairchild alumni, and founded a little company called "Intel" the year before--in Mountain View, CA. Get the idea of why the Santa Clara valley got the name "Silicon Valley").

The AMD 2901 was a hugely popular bipolar bit-slice popular. It could be found in many minicomputers. Intel likewise had the 3000 series bitslice but the 2901 was much easier to work with.

First of all, the MCM/70 was portable, with the option to add a printer. How many portable computers were there in 1973? The choice of the one-line display was about all one could get in a small package back then. For a somewhat later example, consider the Artec word processor of about 1975:
102628783.1.lg.jpg

The photo shows the keyboard of a unit after Artec was acquired by Dictaphone.
Consider early portable computers in general, such as the Radio Shack TRS-100:
102646802p-03-01.jpg
 
Oops, brain got mixed up, I was thinking of ATI (which didn't even exist in the 70s). Apologies!

EDIT: Agreed, that's impressive for 1973 (battery backup, virtual RAM onto the cassettes). IIRC, one of the new things about System/370 was the idea virtual memory.

The Micro Computer Machines MCM/70, the Canadian Holy Grail of Computing History – VintageComputer.ca

vintagecomputer.ca vintagecomputer.ca
There's a note about it fitting under an airplane seat. Could this power off DC?
What was "special" about the tapes? Couldn't be regular audio tapes?
 
Last edited:
You must log in or register to reply here.
Back
Top