Timeline for Why did CPU designers in the 70s prioritize reducing pin count?
Current License: CC BY-SA 4.0
32 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| Jun 18, 2020 at 8:29 | history | edited | CommunityBot |
Commonmark migration
|
|
| May 30, 2019 at 15:11 | comment | added | supercat | The problems of multiplexing could have been much more effectively turned into advantages if more forms of static RAM and EPROM chips included address latches. A memory which latches A0-A7 on the falling edge of /CE could get by with eight fewer connections if wired to a processor that multiplexes those pins. | |
| May 30, 2019 at 7:43 | history | edited | Raffzahn | CC BY-SA 4.0 |
deleted 1 character in body
|
| Apr 11, 2018 at 16:24 | comment | added | supercat | I hadn't realized until I decided to compare it to the 6510 that the 6502 leaves a lot of pins unused. That seems curious, given how many other 40-pin devices seem squeezed for pins. | |
| S Apr 4, 2018 at 16:15 | history | suggested | Peter Mortensen | CC BY-SA 3.0 |
Copy edited (e.g. ref. <https://en.wiktionary.org/wiki/on_the_whole#Prepositional_phrase>).
|
| Apr 4, 2018 at 15:23 | review | Suggested edits | |||
| S Apr 4, 2018 at 16:15 | |||||
| S Apr 3, 2018 at 20:17 | history | suggested | Dennis Williamson | CC BY-SA 3.0 |
remove extraneous apostrophe, make dashes consistent
|
| Apr 3, 2018 at 20:14 | review | Suggested edits | |||
| S Apr 3, 2018 at 20:17 | |||||
| S Apr 3, 2018 at 9:12 | history | suggested | JDługosz | CC BY-SA 3.0 |
And for an
|
| Apr 3, 2018 at 5:05 | review | Suggested edits | |||
| S Apr 3, 2018 at 9:12 | |||||
| Apr 3, 2018 at 0:20 | comment | added | Raffzahn | Gee,@Tommy, there are usable arguments for both ways. In the end it all comes down to what the target design is and what other chips are to be used. Isn't it? | |
| Apr 3, 2018 at 0:13 | comment | added | Tommy | @immibis (and Jules) oh! I see your point now! You've cut down on the number of signals for maybe the 2cm immediately next to the CPU; you’ve had a negligible effect on the board overall and that extra latch/etc is needed. It is entirely clear to me now where my comprehension was failing. Thanks to both for sticking with it through my mental fog! | |
| Apr 2, 2018 at 23:17 | comment | added | Criticizing Israel not allowed | @Tommy All of the pins that you're removing from the CPU are being added to another chip instead - plus some more! So there are no savings in the number of connections. | |
| Apr 2, 2018 at 22:23 | comment | added | Tommy | (I can't seem to find an image of the TI-99/4, but the 4a is in the mid-30s; I struggle to think of any later market entrants prior to 1980 that it's likely I'd be able to find a board picture of) | |
| Apr 2, 2018 at 22:06 | comment | added | Tommy | @Jules just from image searching and visual inspection, I count at least 62 ICs in the TRS-80, 66 in the PET 2001; only the Atari 400 is in the low 20s, getting towards the 1980s level of integration. But it's also hard to support the prospect that everything else would be integrated but the memory latch, so counting it as 1/60th of a machine feels fair. I still don't see how saving 1/60th of the ICs could justify adding more than 50% to the number of traces manually to be wrapped or drawn. | |
| Apr 2, 2018 at 21:47 | comment | added | Jules | @Tommy - the Apple II was a bit of an outlier due to the fact that it implemented bitmapped raster video in discrete logic -- earlier machines typically used a serial terminal for their user interface (with the Apple I being the only exception I can think of, and that was a somewhat simpler monochrome text only design), and later ones tended to use either off-the-shelf CRT controllers or custom ICs for the purpose. | |
| S Apr 2, 2018 at 21:33 | history | suggested | Malady | CC BY-SA 3.0 |
Spelling Corrections
|
| Apr 2, 2018 at 21:08 | review | Suggested edits | |||
| S Apr 2, 2018 at 21:33 | |||||
| S Apr 2, 2018 at 20:10 | history | suggested | Laurel | CC BY-SA 3.0 |
Fixed formatting with real headers
|
| Apr 2, 2018 at 19:41 | review | Suggested edits | |||
| S Apr 2, 2018 at 20:10 | |||||
| Apr 2, 2018 at 17:14 | history | edited | Raffzahn | CC BY-SA 3.0 |
added 1239 characters in body
|
| S Apr 2, 2018 at 16:49 | history | suggested | Toby Speight | CC BY-SA 3.0 |
Spelling and grammar
|
| Apr 2, 2018 at 16:37 | comment | added | Tommy | @Jules I'm not persuaded by that argument, as one extra IC when micros of the late-'70s would often be dozens (an outlier being the original Apple II at 120 chips because it's almost entirely discrete logic) isn't a lot. A reduction in quantity of connections would make life a lot easier to the designer than eliminating a single chip. I could easily see your argument if it were the '80s. | |
| Apr 2, 2018 at 16:36 | review | Suggested edits | |||
| S Apr 2, 2018 at 16:49 | |||||
| Apr 2, 2018 at 16:34 | comment | added | Toby Speight | Not just the production equipment, but other existing tooling (at manufacturer and at customer) - e.g. chip testers, PROM programmers, etc. | |
| S Apr 2, 2018 at 16:23 | history | suggested | Anonymous | CC BY-SA 3.0 |
corrected typos
|
| Apr 2, 2018 at 16:09 | comment | added | Jules | @Tommy - given that to use a chip with multiplexed pins like that you're going to need to add an extra address latch that wouldn't be needed otherwise, and for a processor with a 20-bit address bus that's going to be at least a DIP24 if not larger, I don't think it makes any difference from that perspective, or maybe even makes it worse. | |
| Apr 2, 2018 at 15:55 | review | Suggested edits | |||
| S Apr 2, 2018 at 16:23 | |||||
| Apr 2, 2018 at 15:45 | comment | added | Tommy | I sort of wondered whether it wasn't also because software help in designing boards wouldn't be used by consumer computer companies for quite a while; manually wire wrapping and then drafting a board with 64 signals (such as for a 68000) gives more than 50% more opportunities for error than one with 40. So when you're picking components, you're going to be hesitant. Any opinion on that as a theory? | |
| Apr 2, 2018 at 15:41 | vote | accept | Jules | ||
| Apr 2, 2018 at 15:00 | history | edited | Raffzahn | CC BY-SA 3.0 |
added 655 characters in body
|
| Apr 2, 2018 at 14:51 | history | answered | Raffzahn | CC BY-SA 3.0 |