Now Reading
Motorola’s Pioneering 8-bit 6800 : Origins and Structure

Motorola’s Pioneering 8-bit 6800 : Origins and Structure

2023-12-12 12:53:28

Motorola 6800 Die Shot – Pauli Rautakorpi, CC BY 3.0 <https://creativecommons.org/licenses/by/3.0>, by way of Wikimedia Commons

Why write in regards to the Motorola 6800 (pronounced ‘sixty-eight hundred’)? For the era that grew up with residence computer systems within the late Nineteen Seventies, the 6800 was the 6502’s barely boring uncle. It was utilized in ‘grown-up’ gadgets, like money machines, nevertheless it was slower, costlier and usually much less fascinating than the 6502.

In actuality, the 6800 was a extremely modern design that, simply as a lot as competing merchandise from Intel, represented a significant step ahead in making computing broadly accessible to all.

Intel would ultimately ‘win’ in microprocessors with the x86 structure, and winners write the historical past, however Motorola deserves simply as a lot respect as an early pioneer. That is in all probability why the 6800 got here prime of our poll of which 8-bit design to put in writing about.

The story of the 6800 can be a narrative of how a company can develop modern expertise with monumental potential, solely to self-sabotage and hand the benefit to a competitor. We’ll see how this occurred in Half 2 of the story of the 6800. We’ll additionally uncover that, remarkably, virtually 50 years after its launch you’ll be able to nonetheless purchase the 6800 at the moment!

Share

For contemporary readers the identify Motorola might be most acquainted as a model of cell telephones made by Motorola Mobility, now a subsidiary of Chinese language multinational Lenovo. Cell telephones mix digital and wi-fi electronics and the roots of Motorola Mobility are in Motorola Inc. which, for a lot of a long time, operated at the forefront of each fields.

Motorola Inc, was based as Galvin Manufacturing Company by the brothers Paul and Joseph Galvin in 1929. Early merchandise included a automobile radio which was given the model Motorola, combining ‘Motor’ and ‘ola’ utilized in different audio merchandise such because the ‘Victorola’ phonograph or document participant.  The automobile radios grew to become so profitable that the corporate modified its identify to Motorola in 1947.

By the early Fifties the way forward for electronics lay within the transistor and Motorola constructed its first manufacturing plant for transistors, at a value of $1.5m, in Phoenix, Arizona. Dan Noble, Vice President of Communications and Electronics, wrote enthusiastically and prophetically, in regards to the new gadgets within the firm’s 1955 annual report:

“The phrase ‘revolution’ will not be too robust to use to the change which is able to happen in digital tools of every kind throughout the subsequent 5 years. The event of the transistor and different semi-conductor gadgets will scale back the scale of kit, enhance the reliability of operation, and considerably scale back the operational energy necessities. Not solely will transistors change vacuum tubes in lots of digital equipments, however they’ll make potential the event of recent gadgets for instrumentation, metering, controlling, detection, communications, and computing.”

By the mid Nineteen Sixties, below the management of Lester Hogan, Motorola was making a wide range of transistors, different discrete semiconductors and built-in circuits in Arizona. Hogan would ‘defect’ from Motorola to Fairchild Digicam and Instrument in 1968, to develop into Chairman and CEO. Taking eight of his Motorola group, nicknamed ‘Hogan’s heroes’, with him, Hogan would change Robert Noyce, when Noyce and Gordon Moore left Fairchild to discovered Intel.

The lack of Hogan and far of Motorola’s senior semiconductor group was only one contributor to a turbulent time for the corporate’s semiconductor division. The business was in the course of a transition from Bipolar to MOS built-in circuits. And quite a lot of rivals have been showing, together with the ‘Fairchildren’, corporations based by former Fairchild staff, together with, crucially, Noyce and Moore’s Intel.

Within the midst of all this variation at first of the Nineteen Seventies, it grew to become obvious to Motorola’s, principally new, senior administration group, that quite a lot of corporations have been engaged on the event of microprocessors.

Tom Bennett, who would develop into a key member of Motorola’s microprocessor group, had been launched to the thought within the type of early sketches for what would develop into the primary business microprocessor, the Intel 4004, by Intel’s Ted Hoff. Bennett had been engaged on calculators at Victor Comptometer in Chicago, when he’d been approached by Hoff to see if he could be considering utilizing the proposed Intel chip in new calculator designs.

Bennett joined Motorola in 1971 to attempt to assist them enter the calculator enterprise. It was quickly clear to Bennett, and others, that the microprocessor had a lot wider utility than calculators. It may resolve plenty of different issues that had beforehand wanted customized, complicated and troublesome ‘random logic’ designs. 

Bennett had duty improvement of customized chips for potential exterior Motorola clients. A kind of chips was a customized microprocessor for Olivetti. Motorola gained the design, however couldn’t manufacture the chip within the required PMOS (p-channel Metal Oxide Silicon) course of. As an alternative the design needed to be licensed to competitor Mostek, who have been capable of construct it, and who launched it because the Mostek 5065. Invoice Lattin would later describe Motorola on the time:

…we had a rotating set of administration, and no one ever appeared to stay very lengthy, and no one ever understood the expertise very deeply, and that was an actual downside, as a result of not solely have been implanters costly. You wanted clear rooms. This was a really costly operation. Motorola needed to get in, however they have been, you realize, didn’t fairly perceive what it was gonna take to get in.

Motorola’s incapacity to fabricate the 5065 should have been embarrassing. Lattin, acquired a, usually direct, recruitment pitch from Intel’s Andy Grove, who had taught Lattin at Berkeley, quoted in Inside Intel:

‘Grove had a really abrasive model,’ Lattin recalled. ‘It did not work properly with me… The recruiting model was: Motorola would not have an opportunity – we’ll kick their butts.’

Lattin declined Grove’s supply. The 5065 saga presumably satisfied Motorola’s administration that they wanted to kind out their manufacturing processes. It actually gave members of the Motorola group precious expertise in designing a microprocessor.

However was making a microprocessor definitely worth the appreciable funding wanted? At first Motorola’s advertising and marketing division quoted anticipated gross sales of solely 18,000 microprocessors over 5 years.

However Bennett and colleagues continued to search for gross sales alternatives. Nationwide Information Corp (NDC) offered a proposal that might have required 200,000 microprocessors. NDC was a ‘time-sharing’ firm that noticed the chance to make use of a microprocessor in terminals related to their central system. Nothing got here of the NDC challenge, nevertheless it confirmed the gross sales potential and offered the impetus to begin a brand new challenge to design a microprocessor..

So a group was assembled in Phoenix, Arizona. Bennett would develop into chief architect of the challenge. He was joined by Jeff LaVell and Invoice Lattin. Collectively they used knowledge that LaVell had gathered from speaking to potential clients, to plot a collection of built-in circuit constructing blocks. These might be used to create a spread of techniques, with an emphasis on people who wanted communications functionality. The 6800 could be joined by chips that would supply reminiscence (6801, 6802 and 6803), enter and output capabilities (6820 ‘Peripheral Interface Adapter’ and 6850 ‘Asynchronous Communications Interface Adapter’) and even a modem chip (6860).

John Buchanan who would do circuit design and the structure of the 6800 and the opposite chips. Rod Orgill assisted Buchanan with evaluation of the design and structure. Invoice Mensch labored on the design of the 6820. Chuck Peddle joined the 6800 group in 1973, after the design of the microprocessor was accomplished however labored on the 6820 after which on the advertising and marketing of the designs. The ultimate group remained small, with round seventeen members in complete. 

To construct the 6800 and the opposite family members the group needed to get each the logic design and the fabrication expertise working. On the logic design Bennett would describe the strategy as follows:

My unique plan for it was to principally use a PLA [Programmable logic Array]-type management. However we couldn’t get sufficient pace out of it. So if you happen to regarded on the chip, the highest a part of it’s all ROM, which does the preliminary decode of the directions, after which after that there was a state counter, and that state counter cycled by means of, relying on what number of separate steps it took to get the instruction accomplished.

The structure of the 6800, with ROM on the prime, registers and the arithmetic and logic unit on the backside and management logic between the 2 could be seen in die pictures and within the the ‘Visual 6800’ simulation, that exhibits how alerts within the 6800 change because the 6800 runs (click on on the image beneath to see the 6800 in operation – extremely beneficial!)

Screenshot of Visible 6800 Simulation

On fabrication, the group made issues more durable for themselves by aiming for a design that wanted only a +5 volt enter, in contrast to the +5,-5 and +12 volt inputs wanted by rivals.

Structure of the 6800’s design began on the finish of 1972, nevertheless it took till February 1974 for the primary elements, working other than a number of directions with bugs, to be accessible.

The 6800 was introduced in March 1974, only a month after Intel introduced the competing Intel 8080. Nonetheless, Intel had the 8080 in full manufacturing by April 1974, while the 6800 continued to be stricken by variety of manufacturing points that Motorola took a number of months to resolve. The 6800 would solely enter full manufacturing in November 1974, giving Intel a significant lead of some months.

The 8080’s designer Federico Faggin later stated:

Ultimately, in the course of 1974, the 6800 got here out. That confirmed that our lead that was a yr and a half was reduce down by 9 months. It was a significant blow to me. However nonetheless, we maintained the lead.

… Motorola used a 5 volt course of expertise that was not prepared for prime time, so their chip was very massive and really gradual. So it was not a match for the 8080, and that was really what allowed the 8080 to achieve market acceptance, even if, in my view, the 6800 had a greater structure. It was a extra common structure. It didn’t endure from this patch over the 8008 that the 8080 represented, and it had a greater addressing mode. It was a nicer machine.

At introduction the 6800 was 29mm^2 and will solely run at 1 MHz when in comparison with the 8080’s 20 mm^2 die measurement and ‘as much as 2 MHz’ clock frequency.

Nonetheless, even when the 5 volt course of wasn’t fairly prepared, the simplicity of the design gave Motorola a key benefit.

And the brand new Motorola household of chips was very succesful. When the 6800 launched, the advertising and marketing materials positioned a number of emphasis on the capabilities of the 6800 household of chips (generally known as the M6800 household) and on the communications capabilities of the household.

See Also

The M6800 household is the primary LSI household designed as a coherent modular constructing block strategy to the implementation of microcomputer techniques. From the deceptively highly effective MC6800 Microprocessor to the byte-organized household recollections, to the aptitude increasing peripheral and communications interface adapters, the M6800 household performs collectively as the full product resolution for microcomputer designs.

Motorola Advert for the M6800 household from the April 1975 Version of Electronics journal

One notable function of the M6800 household at launch was the standard of the documentation for the chips. Motorola’s 700 web page handbook for the system even confirmed how one can use the 6800 household to create a whole ‘Level-of-Sale’ terminal.

Structure of a whole ‘transaction terminal’ system utilizing M6800 household chips

So within the 6800 and the M6800 household, Motorola had a critical contender.

And it’s clear from the funding that Motorola’s administration made within the challenge that they understood the significance of the 6800. An early 6800 improvement system was demonstrated to Motorola’s then CEO Bob Galvin, the son of one of many founders of the corporate.

It’s all of the extra shocking then that Motorola’s administration must be careless with the group that designed the 6800. So careless, actually, that their actions would result in the creation of a brand new and highly effective competitor to their new design. A competitor that might change the course of computing historical past. We’ll see how this occurred in Half 2 of the story of the 6800.

In contrast to the Intel 8080, the 6800 wasn’t primarily based on an current microprocessor design. The Motorola design group largely had a ‘clean slate’, throughout the limitations of the expertise of the time, so it’s fascinating to have a look at their decisions.

The 6800 design could be contrasted with the 8080. The Intel processor had plenty of common objective registers (A,B,C,D,E,H,L) a few of which might be grouped to kind 16-bit registers (BC, DE, HL). In contrast the 6800 had simply two 8-bit accumulators (A and B) and a 16 bit index register (IX). In contrast to the 8080, the 6800 had no 16-bit addition or subtraction.

The PDP-8 is commonly quoted because the mannequin for the 6800, nevertheless it’s solely a really unfastened mannequin. The PDP-8 was a 12-bit machine, with a single 12-bit common objective ‘accumulator’.

Wikipedia has some instance 6800 meeting language code which exhibits off the overall model and a number of the weaknesses of the 6800. The pattern code copies a block of reminiscence from one location to a different. The repeated 16-bit hundreds and shops from reminiscence of the supply and vacation spot of the copy (utilizing the index register ix – eg ‘ldx src’ ‘stx src’) is inefficient when in comparison with an structure that might hold two reminiscence addresses in registers on the identical time. The 6800’s strategy contrasts with the relative economic system and effectivity of the identical process on an Intel 8080, which makes use of the 8080’s a number of 16-bit registers.

Reminiscence copy in 6800 meeting

… and in 8080 meeting

Regardless of the decrease clock frequency and smaller set of registers, there have been methods through which the 6800 would outperform the 8080. 

A easy instruction to load an 8-bit fixed into the accumulator (‘MVI A, #FF’ in 8080 meeting language) would take 7 clock cycles within the 8080, whereas the corresponding instruction (“LDA A #FF” in 6800 meeting language) took simply two clock cycles, that means that even at half the ‘clock pace’ the 6800 would execute this instruction virtually twice as rapidly.

The 6800 additionally had some helpful addressing modes. The listed mode would add an 8-bit fixed to the 16-bit index register to kind a brand new 16-bit handle, which might be used as a supply for a load or an arithmetic, logic or evaluate instruction. Thus a single two byte instruction may carry out an ‘add’ to create a brand new reminiscence handle and carry out an arithmetic operation and all in simply 5 clock cycles.

On paper the 6800 might have regarded prefer it was, as Intel’s Faggin stated, ‘very gradual’, however Motorola would argue that it was ‘deceptively highly effective’. The rivalry between 8080 derived and 6800 model designs could be a function of the microcomputer market over the subsequent decade.

Source Link

What's Your Reaction?
Excited
0
Happy
0
In Love
0
Not Sure
0
Silly
0
View Comments (0)

Leave a Reply

Your email address will not be published.

2022 Blinking Robots.
WordPress by Doejo

Scroll To Top