"Motorola 68000 Oral History Panel"

[RedskullDC]

Hi All,

Thought some of you may find this interesting...

http://www.computerhistory.org/colle...alog/102658109

"Panelists:Jack Browne, Murray Goldman, Thomas Gunter, Van Shahan, Billy D. Walker.
Members of the management, design, manufacturing, and marketing teams responsible for Motorola’s 68000 family of microprocessors and peripheral products discuss the evolution of their activities from the 1970s through the 1990s."

Cheers,
Red

[emufan]

about time the Vampire/Apollo team becomes honorary members to that club

[matthey]

Originally Posted by RedskullDC

Hi All,

Thought some of you may find this interesting...

Thanks Red. Yes, interesting video.

Originally Posted by Murray Goldman

Motorola microprocessors went from sales of zero to sales of about 250 million dollars almost overnight ... and it was good margin.

Motorola microprocessors rose from nothing with embedded processors (GM one of the biggest customers) which provided the funding to make the 68000 possible.

Originally Posted by Tom Gunter

None of us had ever designed a computer ... the actual architecture and stuff was defined by the entire staff of teams.

Real teamwork by professionals.

Originally Posted by David House

I can tell you at Intel it [the 68000] was pretty electrifying too... it was terrifying.

The 68000 almost destroyed the x86 competition at introduction. David also called the 68k a superior architecture. Sadly, the x86 won in the long run as Intel did not give up, improved turn around times, gave developers the resources they needed and improved the ISA. It was actually Motorola who threw in the towel on the 68k and moved to PPC which would be beaten badly by x86/x86_64. Too bad we didn't hear about the demise of the 68k. I bet the decision to terminate the 68k and not let it be competitive with the PPC was made by higher ups in suits.

Fragmentation and lack of standardization in the 68k market with many different Unix/Linux flavors and other OSs is mentioned as one of the demises of the 68k. Meanwhile, x86 was united on one OS requiring one standard CPU. Apple is mentioned plenty but Amiga is never mentioned, even when talking of the failures of the Macintosh and the failures of Motorola's gfx chip designs for the 68k. Then again, most people consider the Amiga and the 68k as failures/losers today.

Originally Posted by emufan

about time the Vampire/Apollo team becomes honorary members to that club

No. Combining an old SIMD unit with the integer unit of a 68k is not innovation. Neither is one mans authoritarianism team work. The project lacks vision, future planning and financing. Yes, they are using innovative technology with an FPGA. Yes, they are generating excitement for a few hundred Amiga geeks. Yes, there are some good people involved. Let's not diminish what these original 68k guys did though. They were successful attracting most of the high end processor market and making lots of money. The sheer scale of their work and contributions to processor design is huge compared to the Apollo-core which is unlikely to even be a blip on the radar.

[emufan]

you are right. but one may ask, with all the enthusiasm they had - why did they let the architecture die.
they could have made a clone with a startup somewhere in silicon valley.
the decision of the motorola management looks alot like the commodore demise

[idrougge]

A clone for what market?

[matthey]

Originally Posted by emufan

you are right. but one may ask, with all the enthusiasm they had - why did they let the architecture die.
they could have made a clone with a startup somewhere in silicon valley.
the decision of the motorola management looks alot like the commodore demise

The 68k began with (positive) hype and ended with (negative) hype. The 68k was probably going to lose the processor high end market anyway. RISC looked pretty good when processor speeds looked like they would scale to infinity and compilers could handle all the complexity that could be moved out of hardware. RISC processors are simpler and easier to develop so probably looked good to Motorola too. The 68k was still a good design and a memory miser when memory was not cheap. The 68060 had very good performance for the clock speed compared to any RISC CPU at that time and was power efficient. These attributes made it good for embedded and mid range personal computers where it still had market share. Motorola must have bought their own hype with RISC, panicked when they saw the sales numbers dropping and jumped the fence. There was likely pressure from IBM and Apple who were major customers as they hyped each other into a frenzy and the AIM agreement. Motorola did eventually go back to embedded with the CPU32, DragonBall and ColdFire but they were mostly minimalist and weak attempts well after they had lost the 68k market. This is one of the reasons why I would like to hear from these guys about the history of the 68k end. They dropped the 68k like a hot potato and acted like it must never compete with PPC while Intel treated their ugly duckling like a baby and it grew up into an ugly goose that laid a golden egg for them.

Originally Posted by idrougge

A clone for what market?

There were clones of the 68000 (Hitachi, Signetics, Rockwell) but there were typically few if any enhancements as the 68k slipped lower and lower in the market range. It did find its way into the embedded market, various consoles, Palm devices, etc. but Motorola's enhancements were bolt-ons, there were limited upgrade paths and the support was tepid practically opening the door for ARM (Thumb 2).

[meynaf]

Originally Posted by matthey

This is one of the reasons why I would like to hear from these guys about the history of the 68k end. They dropped the 68k like a hot potato and acted like it must never compete with PPC while Intel treated their ugly duckling like a baby and it grew up into an ugly goose that laid a golden egg for them.

Indeed they're not speaking at all about the 68k end. It sounds like if it has only been an amusing experience for them. Perhaps they just believed in the RISC lies and still believe in them today.

[pandy71]

68K have quite large market - military traditionally used a lot of 68K - they was reliable, relatively easier to program without errors (programmer errors).
Still there is is few active 68K production - radiation hardened for space industry...
Speed is not all, sometimes architecture and code maturity is more beneficial.

[EugeneNine]

This one "Motorola microprocessors went from sales of zero to sales of about 250 million dollars almost overnight ... and it was good margin." isn't quite right either. Motorola made lots of sales with the 6800 series before the 68000 so the 68000 didn't take them from nothing. Though not as much sales as the 6500 the 6800 was IIRC in the trash80 and a lot of embedded boards.

[eXeler0]

Originally Posted by matthey

No. Combining an old SIMD unit with the integer unit of a 68k is not innovation. Neither is one mans authoritarianism team work. The project lacks vision, future planning and financing. Yes, they are using innovative technology with an FPGA. Yes, they are generating excitement for a few hundred Amiga geeks. Yes, there are some good people involved. Let's not diminish what these original 68k guys did though. They were successful attracting most of the high end processor market and making lots of money. The sheer scale of their work and contributions to processor design is huge compared to the Apollo-core which is unlikely to even be a blip on the radar.

@matthey I know your relstionship to the Apollo team is a bit different compared to the rest of us, but even if the paper specs dont impress you, we still need to acknowledge a few things:

First of all, its done in spare time as a hobby. Most ppl spend it on their asses wathing TV :-) Pretty sure the Motorola engineers got payed a lot for their work. :-)

Secondly, talk only gets one so far. What we have here is actually a usable, physical product that respects the 68k legacy and is extended in a highly compatible and usable way and in the end -by far- surpasses the performance/dollar ratio of anything ever released in Amigaland. Thats gotta be worth some recognition. I sure havent seen anyone else deliver that kind of performance at that price. If it was easy, it would have happened a long time ago.


Skickat från min HTC One via Tapatalk

[Mrs Beanbag]

re the end of 68k...

i think it was some amount of wishful thinking. It was clear that the x86 PC was king of Businessland, and despite its obvious shortcomings as a design, Intel had the resources to make it perform. Motorola and others wanted to believe something could beat it, and not unreasonably they believed something radically new was required to do it... logically the arguments for RISC made sense at the time, and really only Intel had their thumb so firmly in the backwards-compatibility pie. So why would they keep 68k? It's easier to see in retrospect that they were too hasty.

As for thw "guys in suits," on the contrary, it was probably the guys in suits who kept Intel on their own track, RISC was always an academic's wet dream (and still is), the business bods probably didn't even understand any of the arguments, they just felt they had to take a risk and put their faith in their engineers.

Did anyone defend the 68k at that time? I'd like to know the answer to that. I mean, apart from me.

[idrougge]

Apart from Apple and Commodore, the previous high-end 68k supporters had already left the 68000 for RISC. SGI had gone for MIPS, Sun for SPARC and HP for PA-RISC. There was obviously something going for RISC.