64E4 - MOS technology 32 bit 68K competitor?

[ChuckT]

Many of those questions are answered in Brian Bagnall's Commodore book.  Also many of those questions have been answered in CBM engineer Bil Herd's videos posted on-line.  For more specifics, why don't you shoot your questions to Bil at http://www.c128.com ?

Happy New Year!
Robert Bernardo
Fresno Commodore User Group
http://videocam.net.au/fcug

Irving Gould got control of the company and Jack Tramiel wasn't happy about it, Jack left and formed Atari.  We know that much.  A book isn't going to tell motives or things that are not discussed.

Irving Gould was a partner and he invested $400,000.  He later got control of the company.

If you wanted to steal a company, what would you do?  You would run it into the ground so the owner couldn't pay his bills.  A company that isn't doing good is worth less.  Right?  The value of a company goes down and you could get a billion dollar company for a steal.  Isn't this what happened to Commodore?  Is the book going to pick up on this?  I haven't read the book but maybe the author was busy on other issues.  Someone observant has to answer and that is why I posed this question to Bil.  He might not be able to go back that far and fully answer the question because it depends on when he joined Commodore.

[RobertB]

That is unfair answer because...

Nobody said anything being fair.  

Happy New Year!
Robert Bernardo
Fresno Commodore User Group
http://videocam.net.au/fcug

[RobertB]

Is the book going to pick up on this?  I haven't read the book...

Enough said.

Happy New Year!
Robert Bernardo
Fresno Commodore User Group
http://videocam.net.au/fcug

[psxphill]

We know why the 1541 was never fixed. The question was just rhetorical in nature.
 
From what I've read, they had about a month.

It only works as a rhetorical question in your favour if you don't know the reason it wasn't fixed. They were going to fix it, they invested money in fixing it. However they found out too late that a mistake in the production meant that they couldn't fix it. Delaying would have allowed another computer to gain market share that they wouldn't have been able to recover.
 
Wherever you read that they only had a month from starting the SID to finishing it was wrong. Both the SID & VIC-II were loosely scoped projects, they don't appear to have been started with any form of deadline. When commodore decided to do the c64 they needed them reasonably functional to demo at the January CES or they wouldn't get any orders. Without orders they wouldn't know whether to manufacture any, a lot of projects were demo'd at CES but if nobody wanted them then they were killed.
 
In answer to your original question. Bob Yannes did the SID on his own and also worked on the C64 motherboard, he left commodore shortly after they were complete.
 
The engineering team at commodore was always very small. At least for the people that achieved anything.
 

They wanted CP/M on a machine so of course you had to use a Z80. I remember the chips they had to use had something to do with the tank at the MOS building was leaking and something about they didn't want to produce one of the chips?

The Z80 in the C-128 was designed in by accident, CP/M was never a design consideration. It was only because someone in marketing had promised 100% compatibility and the C-64's CP/M cartridge worked less in the C-128 than it did in the C-64. Personally I think they could have argued their way out of it as the CP/M cartridge was so rare. Even if it meant designing a new CP/M cartridge that could be exchanged. But Bil chose to put the Z-80 on the motherboard instead, maybe it was cheaper to do it that way? He might have been forced to remove it, except the compatibility hack for the Magic Voice cartridge also relied on it.
 
The MOS building was shut down by the EPA years later due to the leak, I don't remember anything happening around the time of the C128. AFAIK Commodore manufactured all the C-128 & 16bit Amiga chips that they had a license for. For AGA they had to outsource some of the chips because their fab lines were out of date. They redesigned the CIA on a more modern production process for the CD32 as they had closed the fab completely by then, both CIA's are included in the AKIKO chip.
 
http://en.wikipedia.org/wiki/MOS_Technology#GMT_Microelectronics
 

If you wanted to steal a company, what would you do? You would run it into the ground so the owner couldn't pay his bills. A company that isn't doing good is worth less. Right? The value of a company goes down and you could get a billion dollar company for a steal. Isn't this what happened to Commodore?

There doesn't appear to be anything like that going on with commodore. Fraud to drive the stock price up maybe, however Jack was never tied to it. Possibly because the man who was thought to be responsible (C. Powell Morgan) and another key witness died before the fraud was investigated.
 
http://www.commodore.ca/history/company/chronology_portcommodore.htm
 
http://books.google.co.uk/books?id=m4QJke60-oAC&pg=PA114&lpg=PA114&dq=C.+Powell+Morgan+atlantic+acceptance+company&source=bl&ots=XWb7VIzoOr&sig=ex4yT1W5A5dH-FOQwLYgRrRguUk&hl=en&sa=X&ei=wqDkUOS3E4WZ0QWx3oCwBw&ved=0CEwQ6AEwAw#v=onepage&q=C.%20Powell%20Morgan%20atlantic%20acceptance%20company&f=false
 
http://www.commodore.ca/history/people/irving_gould.htm

[psxphill]

They wanted CP/M on a machine so of course you had to use a Z80. I remember the chips they had to use had something to do with the tank at the MOS building was leaking and something about they didn't want to produce one of the chips?

The Z80 in the C-128 was designed in by accident, CP/M was never a design consideration. It was only because someone in marketing had promised 100% compatibility and the C-64's CP/M cartridge worked less in the C-128 than it did in the C-64. Personally I think they could have argued their way out of it as the CP/M cartridge was so rare. Even if it meant designing a new CP/M cartridge that could be exchanged. But Bil chose to put the Z-80 on the motherboard instead, maybe it was cheaper to do it that way? He might have been forced to remove it, except the compatibility hack for the Magic Voice cartridge also relied on it.
 
The MOS building was shut down by the EPA years later due to the leak, I don't remember anything happening around the time of the C128. AFAIK Commodore manufactured all the C-128 & 16bit Amiga chips that they had a license for. For AGA they had to outsource some of the chips because their fab lines were out of date.

[ChuckT]

The Z80 in the C-128 was designed in by accident, CP/M was never a design consideration. It was only because someone in marketing had promised 100% compatibility and the C-64's CP/M cartridge worked less in the C-128 than it did in the C-64. Personally I think they could have argued their way out of it as the CP/M cartridge was so rare. Even if it meant designing a new CP/M cartridge that could be exchanged. But Bil chose to put the Z-80 on the motherboard instead, maybe it was cheaper to do it that way? He might have been forced to remove it, except the compatibility hack for the Magic Voice cartridge also relied on it.

I always thought that they were scared of the popularity of CP/M and wanted a product in case there was competition and I believe the Commodore magazines made mention of it.

I think CP/M was more popular overseas.

[ChuckT]

There doesn't appear to be anything like that going on with commodore. Fraud to drive the stock price up maybe, however Jack was never tied to it. Possibly because the man who was thought to be responsible (C. Powell Morgan) and another key witness died before the fraud was investigated.

There was fraud.  They were a Billion dollar company and they weren't paying their taxes and then the IRS got on them.  That is why they registered in the Bahamas with a P.O. Box.

[ChuckT]

Wherever you read that they only had a month from starting the SID to finishing it was wrong. Both the SID & VIC-II were loosely scoped projects, they don't appear to have been started with any form of deadline. When commodore decided to do the c64 they needed them reasonably functional to demo at the January CES or they wouldn't get any orders. Without orders they wouldn't know whether to manufacture any, a lot of projects were demo'd at CES but if nobody wanted them then they were killed.

That is a recipe for disaster and it is failure on the part of leadership.  You have consumers who have money and who can't travel to the CES show because they work, can't afford a plane ticket and a hotel but they can afford a computer who would buy the product but the computer products aren't made available to them.  Tell me something.  If Commodore came out with a Commodore 128 Slim without CPM and was half the price, do you think people wouldn't buy it if there weren't any orders at CES?

I have a motor and gear factory near me but they will do business with me if I place a 10,000 or 20,000 piece order but since I don't have that much startup capital, they simply won't meet with me so that I could buy 100 so that I could develop a product line for robotics.  Its not that they can't do business but they don't want to do business.

When you do business like that, who needs salesmen?

[ChuckT]

It only works as a rhetorical question in your favour if you don't know the reason it wasn't fixed. They were going to fix it, they invested money in fixing it. However they found out too late that a mistake in the production meant that they couldn't fix it. Delaying would have allowed another computer to gain market share that they wouldn't have been able to recover.

If they waited or if they came out with another revision, would you have bought it?  Would other people have bought it?  And what do you think would have happened to the popularity of Commodore if the 1541 was sped up?  Exactly.

They're doubling the memory on the Raspberry Pi and are people returning their old units?  No.  People are ordering the Raspberry Pi again with double the memory.  It generates more income and makes the product more popular.

Not doing it hurts the company.

Commodore should have had me as manager.  I would have made moves to make them more popular.

[psxphill]

I always thought that they were scared of the popularity of CP/M and wanted a product in case there was competition and I believe the Commodore magazines made mention of it.
 
I think CP/M was more popular overseas.

Bil Herd is on record that he designed it in for the reason I gave. There was nobody at commodore that was working on strategy.
 
CP/M for the Z80 was basically dead way before the C-128 came out. An 8086 would have made more sense as a backup plan.
 

There was fraud. They were a Billion dollar company and they weren't paying their taxes and then the IRS got on them. That is why they registered in the Bahamas with a P.O. Box.

There was no fraud that would drive the cost of the stock down so that it could be bought. They kept having to borrow money to keep going.
 
On the subject of the IRS. 12 years after they were incorporate with the head office in the Bahamas, the IRS decided not to believe them.
 
http://articles.philly.com/1989-01-04/business/26121801_1_commodore-international-tax-bill-commodore-executive
 
In 1990 they started having their stockholder meetings in the Bahamas, which is where Irving Gould lived. So it's likely they stood by this statement.
 
A lot of companies still do stuff like this today and while Starbucks have been shamed into actually paying corporation tax in the UK, there wasn't anything legally that our government could do about it.

[ChuckT]

Bil Herd is on record that he designed it in for the reason I gave. There was nobody at commodore that was working on strategy.
 
CP/M for the Z80 was basically dead way before the C-128 came out. An 8086 would have made more sense as a backup plan.

I'm not questioning Bil.  It doesn't mean that Commodore representatives and Commodore magazines didn't put a spin on it or give their own answer.  Commodore had their own magazine(s).

The Z80 is still used today and was used in some popular video game machines.  Compare the ability of the 6510 to the Z80 and you have your competition.  If you are selling Commodore computers with a 6510 chip then your competitors know there is a market and anyone wanting to bring the competition can do so with a Z80 because it can do as much or more.  The Z80's clock speed today is up to 50MHZ.  The reason you can't speed up the 6502 is because it has multiple clocks so the Z80 would have won out if there was competition on power.

http://en.wikipedia.org/wiki/Zilog_Z80

Quote:

Z80 was often used in coin-operated arcade games,[10] and was commonly used as the main CPU, sound or video coprocessors. Pac-Man arcade games feature a single Z80 as the main CPU.[55][56] Galaxian and arcade games such as King & Balloon and Check Man that use the Namco Galaxian boardset also use a Z80 as the main CPU.[57] Other Namco licensed arcade games such as Galaga and other games that use the Namco Galaga boardset such as Bosconian, Dig Dug, Xevious, and Super Xevious use three Z80 microprocessors running in parallel for the main CPU, graphics, and sound.[58]
It was also found in home video game consoles such as the ColecoVision,[59] Sega Master System[60] and Sega Game Gear video game consoles, as an audio co-processor in the Sega Mega Drive and as an audio controller and co-processor to the Motorola 68000 in the SNK Neo-Geo.
Various scientific and graphing calculators use the Z80, including the Texas Instruments TI-73, TI-81, TI-82, TI-83, TI-83+, TI-84+, TI-85 and TI-86 series.[62]

Quote:
The µPD780C was used in the Sinclair ZX80 and ZX81, original versions of the ZX Spectrum, and several MSX computers, and in musical synthesizers such as Oberheim OB-8 and others. The LH0080 was used in various home computers and personal computers made by Sharp and other Japanese manufacturers, including Sony MSX computers, and a number of computers in the Sharp MZ series.[36]

The Z80 has been made and sold all over the world and is still in use today.

[commodorejohn]

It's a bit nuts to suggest that the decision to add the Z80 was in any way influenced by the fact that, decades down the line, it runs much faster than the 6510 did in the mid-'80s. At the time, the usual clock range for the Z80 was 2-4MHz (I don't know whether they'd got up to 8MHz parts at this point, but if they did they certainly weren't common.) And that's just clock speed, which is a notoriously inaccurate indicator of performance between different architecures. The Z80 is a fairly capable CPU, to be sure, but it takes some 2-5 times more cycles per instruction than the 6502; then again, it can do a bit more per instruction. It mostly comes down to code quality. The reason so many of Commodore's competitors used the Z80 is because Zilog, unlike MOS, wasn't owned by Commodore.

(Also, the 6502 does not use multiple clock signals.)

In any case, psxphill is right; CP/M was pretty nearly dead by 1985, having been supplanted for business use by MS-DOS, and never really catching on with most Z80-based home computers the way it did with the hobbyist micros of the late '70s.

[ChuckT]

It's a bit nuts to suggest that the decision to add the Z80 was in any way influenced by the fact that, decades down the line, it runs much faster than the 6510 did in the mid-'80s. At the time, the usual clock range for the Z80 was 2-4MHz (I don't know whether they'd got up to 8MHz parts at this point, but if they did they certainly weren't common.) And that's just clock speed, which is a notoriously inaccurate indicator of performance between different architecures. The Z80 is a fairly capable CPU, to be sure, but it takes some 2-5 times more cycles per instruction than the 6502; then again, it can do a bit more per instruction. It mostly comes down to code quality. The reason so many of Commodore's competitors used the Z80 is because Zilog, unlike MOS, wasn't owned by Commodore.

(Also, the 6502 does not use multiple clock signals.)

In any case, psxphill is right; CP/M was pretty nearly dead by 1985, having been supplanted for business use by MS-DOS, and never really catching on with most Z80-based home computers the way it did with the hobbyist micros of the late '70s.

{Quote}
Contemporary Z80 processors now have (close to) single-cycle execution rates, thus eliminating the 6502's performance advantage over the Z80.

The 8080/Z80 family offer a larger number of internal registers, distinct I/O and memory address spaces.
{EndQuote}

http://forum.6502.org/viewtopic.php?p=10175#p10175

Previous Memorize Share Next The 6502's two clock signals
The 6502, that classic CPU, is relatively simple in its design as CPUs go, but it has one design quirk which often confuses people studying its architecture: While it has only one clock-input pin (as was usual for CPUs of its time), it has two clock-output pins. Not only are these two clocks made available to outside devices via pins on the 6502 chip package itself, but these two clocks are used internally within the 6502 for various timing and control purposes. Why does the 6502 have two internal clocks, and how are they different from each other?
[EndQuote]

http://lateblt.livejournal.com/88105.html

[commodorejohn]

*facepalm*

Would you care to explain to me what in blazes the performance improvements of contemporary (i.e. modern) Z80 designs has to do with a computer designed in 1985? Hell, even the R800 (in the MSX Turbo-R) wasn't released until ~1990.

And if you'd actually read that second post: yes, the 6502 has two clock outputs. One of which is an inverted version of the other. It only has the one clock, which is why it only has one clock input.

Geez Louise.

[mongo]

If they waited or if they came out with another revision, would you have bought it?  Would other people have bought it?  And what do you think would have happened to the popularity of Commodore if the 1541 was sped up?  Exactly.

They're doubling the memory on the Raspberry Pi and are people returning their old units?  No.  People are ordering the Raspberry Pi again with double the memory.  It generates more income and makes the product more popular.

Not doing it hurts the company.

Commodore should have had me as manager.  I would have made moves to make them more popular.

A Raspberry Pi is $35, a C64 + a 1541 would set you back by about $1000. Do you really think people would buy another one because the floppy drive was faster?

The C64 is the best selling personal computer of all time. Commodore did pretty well without you.

[Hattig]

The Z80 is still used today and was used in some popular video game machines.  Compare the ability of the 6510 to the Z80 and you have your competition.  If you are selling Commodore computers with a 6510 chip then your competitors know there is a market and anyone wanting to bring the competition can do so with a Z80 because it can do as much or more.  The Z80's clock speed today is up to 50MHZ.  The reason you can't speed up the 6502 is because it has multiple clocks so the Z80 would have won out if there was competition on power.

Both the Z80 and the 6502 (and derivatives like the 6510) were decent designs. The Z80 had higher clocks, but a lower instructions per clock. Overall a 4MHz Z80 and a 1-2MHz 6502 performed the same.

The problem was that for a long time, the 6502 stayed at 2MHz max, and additionally the C64 kept on using a 1MHz 6502. The Z80 got a B revision at 6MHz - not that it was used much.

Overall home computers in most of the 80s were in stasis - not continually improving. MOS could have done a new 6502 running at a higher speed, but they didn't for quite some time - and then they lost it all because companies started sourcing them from other sources who did put the work in to improve the design.

Note the C65 was meant to have an up-to ~8MHz 6502 derivative.