What if Woz got what he wanted for a 8MHz version of Apple IIGS

[LisaXL]

We know that the 68000 require 4 clock cycles to access RAM comparing only 2 cycles for 65C816. That makes 65C816 runs roughly twice as fast as 68000 at the same clock speed. I saw it on an interview that Woz originally wanted 8MHz of 65C816 on the IIGS. Won't that mean the Apple IIGS could be as fast as the MAC II that was released six months later? If Apple has released this beast in 1985 when the chip was in full production, the phrase "Apple II forever" can probably have a while new meaning. Perhaps Apple didn't have to go through the trouble of developing Lisa and Mac afterall. They could just focus on improving the Apple IIGS with GUI and higher resolution graphics.

Although history can not be re-written. But if someone at Apple had a vision of focusing on revolving the Apple II line the same way as IBM PC compatibles. We can still be using a form of Apple II today.

 

[olePigeon]

You can get a taste of it with the then-released 8MHz (roughly) accelerators such as the TransWarp GS and ZipGS. I think the upgrades went all the way up to 14MHz at one point. There were aftermarket upgrades up to 20MHz. IIgs got pretty insane.

 

[Cory5412]

It's my understanding that an accelerated IIgs is still meaningfully slower than a Mac for a handful of architectural reasons that you really can't get around unless you drop the 8-bit compatibility or re-architect the entire system so that 8-bit functionality is subservient to the 16-bit platform. (Similar to, say, how the Mac LC IIe LCPDS card worked.)

An Apple-built 16-bit multimedia platform re-architected so it was actually faster would have been real neat.

Idly, one thing I see a lot, and, actually experienced myself in school is that Apple IIgses were, in practice, used as faster/batteries-included IIes. We had a few IIgses at my school when I was a kid and they were used exclusively to run 8-bit edutainment software.

The problem any new computer platform introduced for or bought for k-12 education in the '70s, '80s, and '90s is that a school is typically a self-contained environment where everybody expects to be able to do anything they want on any machine, and, more importantly, where if you've got one or two rooms with 30 computers in them, you can't just sub 5 of those machines out for computers from a completely different platform because usually in K-12 the point of going to the computer lab is that everyone in the room is roughly speaking doing the same work.

In higher education you can get away with it more but higher education was more likely to have people using terminals to connect to big iron machines or to have students using their own machines. so you can get away with a mixed environment much more easily.

Anyway, at that point the question is... do you (Apple in the '80s) have a second machine up your sleeve for K-12 and that's probably why the IIgs ended up architected the way it was so it wasn't yet another attempt to move away from the II, but, more of an iteration -- with all the compromises that entailed.

By way of timelines:
- 1983 - Apple Lisa launches
- 1984 - Apple Mac launches
- 1986 - Apple IIgs launches

so I don't think "what if better IIgs" prevents Apple from doing anything with the 68000 unless you can get Apple a reliable 68516 at like 8-16MHz in like 1980. (That chip was announced/released in 1983 so the answer there is probably no.)

By way of interesting alt-history potential: I've always wondered whether or not a re-architected Apple IIgs would have been a better seller in a home environment where the potential for color and cdrom and faster performance plus integration with bigger/cheaper TVs as displays for some things could have been beneficial.

It would surely have been an interesting time to buy a computer. We had a discussion on here a few years ago about what you'd buy if you had $AMOUNT in 1986. I picked the Mac Plus because hindsight is 20/20, but the Apple IIe sold until like 1993 (outlasting the IIgs) and I think the IIc was around as well, so if you could get what you needed done on one of those you could get a solid productivity computer with a big ecosystem for less than what a IIgs or a Mac cost.

To that point: these platforms are really only as good as the software that runs on them and AppleWorks/HyperCard GS are "okay" -- they don't set the world on fire. They would've been terrible to use on a composite monitor and they're barely tolerable on the purpose-built RGB monitor. It feels like AppleWorks in particular is there to check a box relative to the Amiga for prospective Multimedia Authoring crowd. HyperCard, I haven't gotten into too much but it's got some nice ups over the Mac version that didn't get resolved on the Mac side of things until like 1997. Color is a big one, so I can see using HyperCard to build a CDROM title on the IIgs, but once you start thinking about going there - the costs pile up. (They would've on a Mac too, and, a lot of the story of the IIgs is that people wanted to do things that just weren't cheap/affordable practical yet.)

 

[AndyO]

We can still be using a form of Apple II today.

Oh, I really hope not!

I think the point you're making regarding the potential for performance in the IIGS is an interesting one, but I don't think it was an opportunity missed.

To my mind, the real issue here is that the Apple II line was destined to come to an end anyway, if for no reason other than that in designing for backwards compatibility, the eventual product would turn out to be overly constrained in comparison to new technologies from other manufacturer/vendors.

The Apple II range of systems were excellent in their day, but they came to an end because they had to, else Apple were painting themselves into a corner, and in the context of the rapid development of systems and turnover of businesses of the time, may well not have survived. As such, what Woz wanted was likely and almost invariably extremely clever, but far less good business sense. The Apple II range survived and sold based on the vast amount of software already in circulation, and as long as it remained backwards compatible, new software would have largely been that too. Even though Jobs could have done with a kick up the backside with a major dose of maturity, he was right that the time had come for something new in a new direction.

I have a IIc, in fact two of them, but even by the time that launched, it was creaking under the millstone of compatibility with much earlier models in the range.

 

[Unknown_K]

There is not a lot of software written for the IIgs so it would have died fast anyway. The Apple II had quite a big software library which is why they kept selling.

 

[LisaXL]

I think if IIGS was released in1985 with 8mhz CPU. Things could be different. In 1986, Atari ST and Amiga had already established strong foothold. In addition, the under powered IIGS had very slow performance under GUI. I think more dedicated IIGS apps could be developed if it was released one year earlier and had faster CPU.

 

[Unknown_K]

The Amiga and Atari ST were cheaper and had a larger game library which is what home users wanted.

 

[LisaXL]

I am really surprised that the number of Apple II family ever sold was just 6 million over the lifespan of 16 years. Amiga sold just as much during its life span. But obviously Apple gets much more profit from Apple II line than Commodore through its Amiga and C64. It is probably primary due to the high price of Apple's computers which give much greater margin on each unit sold.

 

[bibilit]

I am really surprised that the number of Apple II family ever sold was just 6 million over the lifespan of 16 years. Amiga sold just as much during its life span. But obviously Apple gets much more profit from Apple II line than Commodore through its Amiga and C64. It is probably primary due to the high price of Apple's computers which give much greater margin on each unit sold.

The Apple II line was a cash cow for many years. The Macintosh line was not very successful at the beginning, the lack of software and memory for the 128 was part of the problem... and the Lisa was a failure. Both had cost a fortune in R & D.

Without the A II series, the company could have been in serious trouble pretty soon.

 

[Trash80toHP_Mini]

68020

 

[Gorgonops]

Won't that mean the Apple IIGS could be as fast as the MAC II that was released six months later?

Per Trash's numeric comment above, the Macintosh II used the 68020, not the 68000, at 16mhz, so, no, an 8mhz 65816 would not be faster than the Mac II. More importantly, a faster IIgs, unless you made a raft of other changes, would still not be a directly comparable machine to the Macintosh in a number of other important respects. It's swell and all that the IIgs has color, but it only has a 200 line vertical resolution and very non-square pixels; it doesn't even have the eye-killing option of 400 line interlaced like the Amiga. (Even the Atari ST offered a 640x400 monochrome option for "productivity applications", which was at least comparable to if not better than the plain Mac's 512x342.) The IIgs as-built also had some severe architectural limitations, such as the video memory only being accessible via a 1mhz channel, that would need drastic changes to scale it up into something comparable to a Macintosh II.

To go even deeper: Even if you take it as read that the 65816 has some performance-per-clock advantages over the 68000 there's still the fact that it's a *completely* 16 bit architecture implemented with a multiplexed 8-bit data bus; the memory layout is essentially segmented just like an 8086 with native addressing based on 16 bit quantities, so it's a far more awkward than the 68000 family to deal with large memory structures. Optimistically the 65816 is more comparable to Intel's 80286 than the 68000, and an interesting thing you'll discover if you delve into comparisons between the 68000 and the 80286 is the '286 will usually kick the 68000's butt on raw integer calculation performance, the 68000 actually doesn't perform that much better than the 8086, but on problems that benefit from large in-memory datasets the 68000 can smoke it. People were getting serious about thinking the future of personal computing was going to be 32 bit addressing, and the 65816 was a dead-end on that front. Giving the IIgs a high-resolution display with more than 32K of VRAM would have right there exposed the serious weaknesses of the platform.

I'm not saying Apple did the "right thing" by hobbling the IIgs' CPU speed quite as much as they did (although in their defense it sounds like they were having yield issues during the development phase and there are other aspects of the architecture that also held them back), but I also totally don't buy into this narrative that the IIgs could have "replaced" the Macintosh. IBM's PS/2 line came out less than six months after the IIgs and brought 640x480 square-pixel graphics resolution to "everyone", even at the low end with the MCGA-equipped Model 30, and anyone who's used either would have no problem telling you that productivity applications were immensely better on that than the IIgs' CGA resolution. The IIgs being faster alone would have done nothing to change that.

 

[Trash80toHP_Mini]

Thanks for explaining so well what I could never have managed, Eudi. Indeed, 68020 was released in 1984, the same year as the Macintosh, two full years before the IIgs was released.

To flesh out Cory's timeline:

- 1983 - Apple Lisa launches
- 1984 - Apple Mac launches
- ________68020 released
- 1986 - January: Mac Plus introduced
_______ - September - Apple IIgs launched
_______ - Apple dev team renegades form Radius
- 1987 - Six months after IIgs release: Mac II and SE are launched with far superior NuBus and PDS slots
- 1990 - 68020 LC/IIe Card combo continues support of massive 8bit K-12 software base and IIgs is canned.

I'd say it's likely the IIgs was relegated to Color offering stopgap status during its development with the LC/IIe Card concept penciled into the marketing plan wish list very early on, if not before IIgs release.

PageMaker was in development specifically for the Mac during 128K development. Radius first product released: Full Page Display for compositing. Second product release: Radius_16 a 68020/FPU accelerator with Cache and the power to really run PageMaker on the Plus.

For all practical purposes IRL (outside the hermetically sealed K-12 niche) the IIgs was DOA the way I see things. The DTP revolution killed off the Apple II series and saved the Compact Mac line until the Color QuickDraw Mac II shipped.

 

[LisaXL]

Thank Gorgonops and everyone's analysis. I fully agree and respect your opinions. But I cannot agree with 65C816 is a dead end. Just like IBM PC started with 8 bit 8088 and look where it is today. I am sure if Apple started IIGS earlier and put proper marketing to it to become successful. The 6502 architecture could have evoled into 32 bit and even 64 bit if there is demand for it.

But of course Apple wanted to focus on newer products, so they have generally gave up on Apple II long ago. They went through a string of failures like Apple III, Lisa, and the original Mac 128 until Desktop Publishing saved the Mac. Had it not for Apple II continued success, Apple would have gone under long time ago. I just wonder what would happen if Apple focusing on improving the Apple II architecture like the IBM PC instead of developing new platforms. Don't get me wrong, the classic Mac is still my favourite OS of all time. I owned several Mac and Powerbooks along the way. Thanks

 

[Trash80toHP_Mini]

32bit architecture of Macintosh II NuBus Cards and I/O blew 16bit ISA out of the water at its release along with anything with a lesser bus. PC didn't catch up until proprietary 32bit Local Bus video was implementations appeared in the 486 era, no?

Micro Channel doesn't count. :-b

 

[Gorgonops]

Thank Gorgonops and everyone's analysis. I fully agree and respect your opinions. But I cannot agree with 65C816 is a dead end

Western Design Center was a small specialist design firm that completely concentrated on embedded systems. There was *nothing* like a serious roadmap for evolving the 65816 into a 32 bit architecture, and the 65816 was already a quirky obsolete-feeling oddball the day it was introduced. There is also the fact that the 65816 itself shipped about two years late, and during the course of its development Apple had repeated fights with the company because the first versions of the CPU had incompatibilities specifically with hardware like Apple’s disk controllers. The companies were actually on pretty poor terms with each other by the time the IIgs design was mostly complete, it isn’t surprising in the slightest that Apple gave up on making any successors to it. Betting the company on WDC delivering a CPU actually capable of meaningfully competing with Motorola, Intel, or frankly any other option (RISC was starting to become a thing…) would have been flat out stupid.

 

[Unknown_K]

32bit architecture of Macintosh II NuBus Cards and I/O blew 16bit ISA out of the water at its release along with anything with a lesser bus. PC didn't catch up until proprietary 32bit Local Bus video was implementations appeared in the 486 era, no?

Micro Channel doesn't count. :-b

First EISA system shipped October 89 (Compaq Deskpro 486 and SystemPro).

 

[Unknown_K]

Apple charged insane amount of money for their computers and nobody is going to keep paying that for anything Apple II in the late 1980's when faster and faster PC clones were dropping in price daily.

 

[Trash80toHP_Mini]

Good point, the Mac suffered as much.

First EISA system shipped October 89 (Compaq Deskpro 486 and SystemPro).

Forgot that one, but EISA wound up in the Server niche.

IIRC the first mainstream 32bit implementation for the PC was the VESA Local Bus of 1992 with its (3per system?) slot count limitations. It came long after Apple's implementation of the severely slot limited PDS (one per system or two if you count an accelerator) originating in the Killy Klip 68000 era, officially blessed by Apple in the SE of 1987.

16bit architecture doomed the IIgs along with the 8bit road kill it left behind.

 

[AndyO]

Had it not for Apple II continued success, Apple would have gone under long time ago.

There is really no arguing with this point, because the Apple II was a cash-cow for the company for years. It sold extremely well because there was so much software available for it. However, the marketplace for systems was evolving rapidly by the mid-80s, and the Apple II was not, nor could it be made to be an evolutionary product. Successful computer manufacturers were going out of business because they weren't keeping up with rapid development cycles, so ultimately, the fact Apple sold so many of the II product range and used those revenues to fund new products was really just how it had to be - and how most successful businesses do it.

Sadly too, thinking that a better and faster processor could have made any of the Apple II variants into a contender that could compete and outclass emerging technologies from other manufacturers is rather like thinking if you simply put a bigger engine in a Trabant, you could compete with Ferraris at Le Mans.

 

[LisaXL]

Western Design Center was a small specialist design firm that completely concentrated on embedded systems. There was *nothing* like a serious roadmap for evolving the 65816 into a 32 bit architecture, and the 65816 was already a quirky obsolete-feeling oddball the day it was introduced. There is also the fact that the 65816 itself shipped about two years late, and during the course of its development Apple had repeated fights with the company because the first versions of the CPU had incompatibilities specifically with hardware like Apple’s disk controllers. The companies were actually on pretty poor terms with each other by the time the IIgs design was mostly complete, it isn’t surprising in the slightest that Apple gave up on making any successors to it. Betting the company on WDC delivering a CPU actually capable of meaningfully competing with Motorola, Intel, or frankly any other option (RISC was starting to become a thing…) would have been flat out stupid.

Thanks for the insight, now it's clear that it was a wise decision for Apple to go with 68000 qnd Mac