Eight More 8-bit Era Microprocessors (2024)

The Signetics 2650 found its way into a bunch of arcade games by lieu of British company Century Electronics. I have a conversion kit of theirs installed into a Donkey Kong Jr. board, which outright replaces DK Jr.'s Z80 with a daughterboard containing the CPU. Always wondered why they chose that in particular, it's not a very common chip, and just using the Z80 that was there and replacing the ROMs was the more common option for conversion kits like that.

I did some assembly programming on the Fairchild F8 mentioned in the prequel article. Quaintest feature: Doing a “long” jump (more than 127 bytes away) would cause the accumulator register to be clobbered. Presumably, there was nowhere else to store the high (low?) order address byte routing things around to the PC register. This was also a problem for the debugger (in ROM on the development system), since continuing from a breakpoint necessitated a long jump, so it couldn’t restore the accumulator. So, the debugger would just simulate instructions until it hit a jump, which it could then jump to. Or something like that. Fairchild provided a listing of the source to the debugger / emulator, and the line that simulated messing up the accumulator during single-stepping was commented “The F8 Touch!” It made an impression 50 years ago.

The RCA-1802 was used in the COSMAC Elf computer which was described as a hobbyist project to build in a series of articles in Popular Electronics 50 years ago. The Elf may be obscure but one thing developed on it (or its successor, the COSMAC VIP), CHIP-8, lives on -- it was (by some definitions) the first "fantasy console" like Pico-8 and TIC-80 today -- a virtual machine designed for writing action games.

https://en.wikipedia.org/wiki/COSMAC_Elf https://en.wikipedia.org/wiki/CHIP-8

It's easy to forget how much innovation divergence was happening in the early 70s. Up till the late nineties we speculated that ISAs other than x86 (spec. RISC designs) would win in the end.

Imagine these 8 bitters were mostly hardwired, with less than a million transistors.

It's easy to forget how much innovation divergence was happening in the early 70s. Up till the late nineties we speculated that ISAs other than x86 (spec. RISC designs) would win in the end.

Imagine these 8 bitters were mostly hardwired, with less than a million transistors.

> ISAs other than x86 (spec. RISC designs) would win in the end

They're correct: ARM won the mobile space, won Apple, and is very gradually seeping into wider availability for other operating systems.

Never mind the majority of raw FLOPS these days are almost certainly going through GPU architectures.

It depends on how you want to treat the internal (not seen by user or compiler) decoding that both AMD and Intel are doing; some would argue that these CPUs are RISC chips that decode x86 and x64 instructions into their own RISC implementation.

Way less than a million! I believe usually in the order of thousands.

A 486 already had over a million transistors. These are in the thousands.

https://en.wikipedia.org/wiki/COSMAC_Elf https://en.wikipedia.org/wiki/CHIP-8

Oh yes, I cut my teeth on a Finnish 1802 based TELMAC in the 1970s. It all of 2 kb of RAM (in 16 chips manually soldered on the board), and room for another 2kb if anyone would need so much, as it said in the instructions. I might almost be able to reconstruct the instruction set now, it was so deep ingrained in my brain. Once I had nightmares directly in hex. Sold my first pieces of code on that CPU, a 2kb long "monitor" rom, maybe a bit like a BIOS in today's terms. Added .5 kb of new features and optimized it so it could still fit in that 2kb EPROM.

Zaccaria pinballs I think used it pretty heavily as well. Again, no clear explanation there either.

Given the era, I wouldn't be surprised if it came down to "someone on the development team liked that part". (Or "someone in the purchasing department got a bunch of them really cheap".)

The 2650 was one of the first microprocessors on the market you could buy, it was shockingly simple to design, and had some dedicated I/O lines. Look up the reference schematic and compare to what a 6502 or Z80 needs. It was used in a lot of early embedded systems. It makes sense in that light

> ISAs other than x86 (spec. RISC designs) would win in the end

They're correct: ARM won the mobile space, won Apple, and is very gradually seeping into wider availability for other operating systems.

Never mind the majority of raw FLOPS these days are almost certainly going through GPU architectures.

Way less than a million! I believe usually in the order of thousands.

Indeed. The 6502 had 3510 enhancement transistors and 1018 depletion transistors for a total of 4528...

The 8-bit monolithic CPUs of the seventies all had a few thousands transistors.

The first generation of true 16-bit CPUs, i.e. Intel 8086, Motorola MC68000 and Zilog Z8000, had almost an order of magnitude more transistors, i.e. in the range of 15000 transistors to 50000 transistors.

The first true 32-bit CPUs, like the National 32000 series, Motorola MC68020 and Intel 80386, had a few hundred thousand transistors.

By the end of the eighties, the second generation of 32-bit CPUs reached 1 million transistors.

The Motorola 68000, a great CPU with 32-bit operands, was initially implemented with 68,000 transistors.

The model number was decided long before the transistor-level design was finalized.

Given the era, I wouldn't be surprised if it came down to "someone on the development team liked that part". (Or "someone in the purchasing department got a bunch of them really cheap".)

Zaccaria pinballs I think used it pretty heavily as well. Again, no clear explanation there either.

The 8-bit monolithic CPUs of the seventies all had a few thousands transistors.

The first true 32-bit CPUs, like the National 32000 series, Motorola MC68020 and Intel 80386, had a few hundred thousand transistors.

By the end of the eighties, the second generation of 32-bit CPUs reached 1 million transistors.

Indeed. The 6502 had 3510 enhancement transistors and 1018 depletion transistors for a total of 4528...

With the layout fitting entirely on a large sheet of paper: https://archive.archaeology.org/1107/features/mos_technology...

The Motorola 68000, a great CPU with 32-bit operands, was initially implemented with 68,000 transistors.

The model number was decided long before the transistor-level design was finalized.

A 486 already had over a million transistors. These are in the thousands.

I think there were actually 68000 transistor positions. In the ROMs and PLAs not every potential transistor is populated but the missing ones were counted as well. But the number of actual transistors is only slightly smaller so it doesn't really matter.

With the layout fitting entirely on a large sheet of paper: https://archive.archaeology.org/1107/features/mos_technology...

I do not remember the exact number of transistors in MC 68000, but I think that it was less than 40000, so not just slightly smaller. In any case, it had more than twice as many transistors as Zilog Z8000, which was super-optimized for size (a very bad decision of Zilog, which lead to a too long time-to-market and to many initial bugs), and slightly more than 4/3 times as many transistors as Intel 8086.

The 68000 transistors number claimed by the Motorola marketing was close to what you get by dividing the die area to the area of one transistor, so it did not correspond to actual transistor positions.

The MC68000 die had large areas occupied with microprogram ROMs, and there as you say only a part of the array of transistors are active, depending on the stored bits. Nonetheless, a significant part of the die was occupied with random logic, where all the physical transistors are used and a part of the area does not have any transistors.

In The Virtues Of The 8-bit Era : Eight Iconic Processor Designs we looked at eight of the most important microprocessor designs of the 1970s and early 1980s. This time we’ll look at more 8-bit designs from that era, that didn’t quite make it into that list.

The rules are broadly as before. The designs have to be for microprocessors and not microcontrollers. They don’t have to have made it into production though.

Most of these designs didn't have a big impact, but just because they weren’t commercially successful doesn’t mean that they weren't interesting. And sometimes we can learn as much from market failure as from success.

So here are some lesser known ‘8-bit era’ microprocessors.

The first entry in the list comes from Texas Instruments. The team that created the first microcontrollers, led by Gary Boone, had earlier worked on a custom design for the Control Terminal Corporation’s ‘Datapoint 2200’ terminal which led to the creation of the TMX-1795 microprocessor. A little later, Intel created the 8008 using the same instruction set architecture.

The TMX-1795 is one of a number of designs that can reasonably be claimed to be the first microprocessor. Ken Sheriff has written about the 1795 and the contesting claims here. It’s a great blog post, and I’m not going to spoil Ken’s conclusions!

The TMX-1795 didn't even make it into production, but it did make an impact decades later. The design was used in the patent lawsuit between Gilbert Hyatt and TI.

The TI laptop has been modified to use the TMX-1795 for all operations. Everything displayed is being executed on a 1971 vintage 8-bit microprocessor. This demo unit was built in 1993 to be used as a jury exhibit in a patent lawsuit, although it was never used for that purpose. Its first public demonstration occurred a few days later, at the 1996 Microprocessor Forum, during the 25th Anniversary of the Microprocessor celebration.

We can see the TMX-1795 booting up in this Computer History Museum video.

The Intel 8008 did make it into production, and the architecture evolved over the coming years into the dominant x86 architecture that we all use today.

TMX-1795 Key Features

Pioneering 8-bit design;
Helped to overturn Hyatt’s microprocessor patent;
A huge missed opportunity for TI?

We’ve encountered the Mostek 5065 already. In Motorola’s Pioneering 6800 : Origins and Architecture, Motorola struggled to manufacture their first microprocessor design, originally intended for use in Olivetti calculators. So they licensed it to Texas rival Mostek, where it became the Mostek 5065.

I think it must have been clear to potential users that the 5065 didn’t have much of a future. Motorola moved on to the 6800 and Mostek showed no interest in updating the design, possibly because the licensing didn't allow them.

The Mk 5065 was a dead-end, so why would you use it?

The Mk 5065 is interesting as it represents their first thoughts of the team that created the Motorola 6800 and the MOS Technology 6502. Some of the features that would make the later designs distinctive, appeared in the 5065 first. These included, for example, using page 1 of memory for the ‘stack’.

Mk 5065 Key features:

Page 1 of memory used as stack page (as in the 6502);
Fast interrupts, using three sets of registers.

Intel 8085 - By Konstantin Lanzet - CPU Collection Konstantin Lanzet, CC BY-SA 3.0, https://commons.wikimedia.org/w/index.php?curid=5791162

The Intel 8085 was Intel’s successor to the groundbreaking 8080. The 8080 had been the CPU that kick-started the whole personal computer revolution.

Federico Faggin, who had led the design of the 8080, had, after falling out with Andy Grove, left Intel to found Zilog. There he led the development of the Z80 which introduced multiple, major improvements on the 8080. These included IX and IY index registers, a complete set of alternate registers, DRAM refresh built in, lots of extra instructions and more.

The 8085, by contrast was a much more modest upgrade. Compared to the 8080 it incorporated a couple of support chips and upped the clock speed to 3,5 or 6 MHz. The 8085 wasn't a bad design. It was just very underwhelming when compared to Faggin’s Zilog Z80 and lost out to the Z80 in the market.

Intel 8085 Key features:

Not much, really, when compared to the Z80, and that was the problem.

The Signetics 2650 was, in contrast to many other designs of the era, designed to compete with minicomputers, and was modelled on the IBM 1130 minicomputer.

It was competitive with the market leaders, including the Intel 8008, when it was first designed in 1972, but didn't make it into production until 1975. By then, the Motorola 6800 and Intel 8080 outclassed the 2650 which had significant limitations, including memory that was divided into 8k segments.

Signetics 2650 Key Features

Vectored interrupts
On-chip call stack
Limited to 32k memory, split into 8k segments

RCA 1802 Die Shot - By Pauli Rautakorpi - Own work, CC BY 3.0, https://commons.wikimedia.org/w/index.php?curid=39374497

In many ways the RCA 1802 was ahead of its time. It was the first CMOS microprocessor, and its simple instruction set was quite RISC-like, almost a decade before the RISC concept became popular. However, this simplicity also meant that lots of common features, such as conditional branching and subroutine calls, were omitted. Perhaps this limited its popular adoption in an era when processors were marketed based on their instruction set features.

The 1802 was used in some video game consoles, though, and was used in a number of spacecraft because (according to CPU Shack):

… a version of the 1802 used silicon on sapphire (SOS) technology, which leads to radiation and static resistance, ideal for space operation.

RCA 1802 Key Features

First low power CMOS microprocessor;
Extreme simplicity;
Used in Voyager, Viking and Galileo Space probes.

After the paywall the final three of our ‘8-bit’ era designs.

Possibly the least well known of the designs on this list, Wikipedia says that it would be considered a microcontroller, but I think that the lack of onboard ROM means that it was really a microprocessor.

Low cost was its selling point with an inexpensive 28-pin package. However, that limited it to 4k of address space.

Electronic Arrays had problems making the design and quickly failed, being sold to NEC in 1978.

Electronic Arrays 9002 Key Features

64 bytes of on-chip RAM;
Low-cost 28 pin package
Limited to 4k memory.

Intersil 6100 - By Pauli Rautakorpi - Own work, CC BY 3.0, https://commons.wikimedia.org/w/index.php?curid=30674996

On the surface the Intersil 6100 had a lot going for it. It was based on, and largely compatible with, the popular DEC PDP-8 minicomputer architecture. It was built using low power CMOS technology. DEC even used it in their own ‘DECMate’ terminal / personal computer product.

DECMate - By The original uploader was Frotz at English Wikipedia. - Transferred from en.wikipedia to Commons., CC BY-SA 3.0, https://commons.wikimedia.org/w/index.php?curid=53859617

Byte Magazine (03/76) : The FIRST [AND LAST?] OF THE 12 BIT MICROS?

But the 6100 never made an impact; perhaps 12-bits was just too unfamiliar in an era dominated by 8-bit designs. The PDP-8 compatibility didn't help much as the microcomputer market was so new and most software was created from scratch.

Intersil 6100 Key Features :

Low power CMOS;
PDP-8 compatibility;
12-bits.

We’ll finish where we started, with Texas Instruments, and the TMS 9900.

The inclusion of the TMS 9900 in this list may be controversial. The TMS 9900 was marketed as a 16-bit processor, but it appeared in the 8-bit era and really competed with 8-bit designs. Despite those extra eight bits, its performance was typically worse than its 8-bit competitors.

The TMS 9900 was actually a single chip shrink of the TI-990 minicomputer and part of TI’s strategy to build a series of designs around a single architecture.

With a highly successful calculator business, it was natural for TI to want to establish itself in the growing home / personal computer market. So it used the TMS 9900 to create the TI 99/4.

TI 99/4 - By Tocchet22, CC BY-SA 4.0, https://commons.wikimedia.org/w/index.php?curid=87545215

The TI99/4, and its successor the TI 99/4A were simultaneously, a huge success and a massive commercial failure. Quoting Wikipedia:

By late 1982, TI was dominating the U.S. home computer market, shipping 5,000 computers a day from their factory in Lubbock, Texas. By 1983, the 99/4A was selling at a loss for under US$100. Even with the increased user base created by the heavy discounts, Texas Instruments lost US$330 million in the third quarter of 1983 and announced the discontinuation of the TI-99/4A in October 1983. Production ended in March 1984.

On the surface TI’s strategy for the TMS 9900 was attractive. TI had the brand, the distribution and the financial resources to make it a success. But the implementation of the design was problematic and TI made a huge mess of the whole project.

There are so many lessons to be drawn from the TMS 9900 and TI 99/4 that they really deserve a follow-up post. Watch this space!

TMS 9900 Key Features

16 bits;
Slow!

Ken Sheriff on the origins and design of the 1795. Great reading as usual!

http://www.righto.com/2015/05/the-texas-instruments-tmx-1795-first.html

A fantastic, detailed exposition of the story of the TMS 9900 from IEEE Spectrum.

https://spectrum.ieee.org/the-inside-story-of-texas-instruments-biggest-blunder-the-tms9900-microprocessor

Mk-5065 Datasheet at Bitsavers

http://www.bitsavers.org/components/mostek/_dataBooks/1974_Mostek_Integrated_Circuit_Guide.pdf

Hacker Times