A1200.net + Vampire standalone = Powerful new 68K Amiga?

[wXR]

Perhaps this is a demonstration of how pathetically obsessed I am with all things Amiga as of late, but I actually had this dream last night:

I received a new "Amiga 1200". I rolled it out of the box, plugged in the 50/60 110-240 universal power supply, and connected one end of an HDMI cable to the Amiga, and one end to a 16:9 flat-panel. Flipped it on, booted it up almost instantly. I ran legacy Amiga apps using custom chips, and RTG stuff alike. All ran perfectly under an enhanced AROS68K.

Guys, can we make this vision a reality? We have A1200.net producing cases, keycaps, and soon the keyboard mechanisms themselves. We have the Apollo team opening up the SAGA core, and some real movement in the world of AROS68K. Why can't we just put all of these together and make a new "classic" Amiga? It seems like the opportunity is just sitting there.

[Samurai_Crow]

Depending on the form factor of the 10x12cm motherboard on the stand alone vampire, yes.

[wawa]

Originally Posted by wXR

Perhaps this is a demonstration of how pathetically obsessed I am

calm down a bit. not that it is a straw fire.

[Kelv]

Wouldn't it be awesome though.
My dream involved a Pci-e version of the Siamese board by the Apollo guys + a pico Itx board with Aros gluing it all together. Yes I know know emulation is more than capable, but I lusted after one of those things ever since it appeared in one of the Amiga mags way back when.

This really does seem like exciting times to be getting back into Amiga's for me... The enthusiasm being shown is inspiring, new open minded hardware being developed, tool chains which may lower the entry level requirements .. Wow.. keep it up guys... please

[matthey]

Originally Posted by wXR

Guys, can we make this vision a reality? We have A1200.net producing cases, keycaps, and soon the keyboard mechanisms themselves. We have the Apollo team opening up the SAGA core, and some real movement in the world of AROS68K. Why can't we just put all of these together and make a new "classic" Amiga? It seems like the opportunity is just sitting there.

Fixing the AmigaOS situation is just the first obstacle. I have been saying for a long time that it doesn't get you very far unless you have a plan to fix later obstacles. It is new affordable Amiga hardware which would drive demand and development of the AmigaOS and software.

1) Make AmigaOS/AROS available, usable and upgradable (restart 68k AmigaOS development or enhance 68k AROS).
2) Use FPGA hardware to make the 68k CPU and custom hardware available, usable and upgradable.
3) Make a 68k ASIC and use an FPGA for the custom hardware.
4) Make a 68k+custom hardware single chip Amiga SoC ASIC.

This was my vision several years ago and I tried to make it happen. There were other important obstacles which needed addressing.

1) 68k compiler support and code generation quality needs to improve
2) An FPGA 68k CPU needs to be fairly optimal, advanced, organized and upgradable.
3) 68k and AmigaOS standards need to be created (Development and Standards Committees)
4) money needs to be raised and partners found
5) products need to be designed and marketed

I worked on 1 above choosing to improve vbcc. I was partially successful in improving 68k compiler support but was mostly unsuccessful at improving code generation quality significantly. For 2 and 3 above, I helped with the Apollo Core guys giving many suggestions for improvements, programmed some 68k statistics gathering software and benchmarks and documented team proposals for an enhanced ISA. I consider my Apollo work a complete failure though as Gunnar decided to hyper-optimize for an FPGA and abandoned the enhanced ISA completely. As far as 4 above, I was already talking to some investors and doing some investigation for them which I myself lost contact after I decided Gunnar was unreasonable. I made first contact with a potential embedded and ASIC partner which went pretty well but there were too many loose ends (uncertainty is risk) to get very far. I never really made it to the marketing part but suggested some ideas on forums and have talked to Matthew@A-Eon suggesting restarting 68k AmigaOS development and improving APIs to be more compatible to AmigaOS 4 to ease development and create a larger market. I suppose I failed here also. Amiga has a way of turning dreams into nightmares. You would save a lot of time and money if you forgot any Amiga visions you had. Sorry, that is the reality of the current Amiga situation.

[Seiya]

many peopole return to use Amiga for a simple retro gaming interest and only in this way amiga could continue to be watched for many years.
Any other project interest only current users..

[matthey]

Originally Posted by Seiya

many people return to use Amiga for a simple retro gaming interest and only in this way amiga could continue to be watched for many years. Any other project interest only current users..

It is important to attract people outside of the current Amiga community. Here is a list of users and uses which could be targeted.

2) Use FPGA hardware to make the 68k CPU and custom hardware available, usable and upgradable.
o previous Amiga users
o retro 68k, 32 bit, 16 bit and 8 bit gamers
o low end GP computing, gaming and console like
o hobbyist and embedded uses needing a large FPGA
o education including FPGA and easy 68k programming

3) Make a 68k ASIC and use an FPGA for the custom hardware.
o previous Amiga users
o retro 68k, 16 bit and 8 bit gamers (limit for an "affordable" FPGA)
o semi-modern GP computing, gaming and console like
o hobbyist and embedded uses needing an FPGA or 68k CPU
o education including FPGA and easy 68k programming

4) Make a 68k+custom hardware single chip Amiga SoC ASIC.
o previous Amiga users
o retro 68k and Amiga gamers
o modern low to mid end GP computing, gaming and console like
o advanced hobbyist and embedded uses
o education including easy 68k programming

The price is very important as lower prices are attractive to a much larger market. Stage 2 would not attract many users besides previous Amiga users and high class retro gamers. Each stage of development could reduce the price per unit but has more up front costs. The target market changes some with each stage. Lower stages would still have some advantages over higher stages. This is a natural progression but higher level stages may not be economically viable depending on many factors which would need evaluation. I do believe stage 3 is feasible now with some risk adverse (perhaps illogical) investing which I do believe is available currently under the right conditions. Stage 1, active AmigaOS 68k development is a big loose end and a major impediment to investment and later development.

[wXR]

@matthey

I don't really get your point about obstacles. A Vampire standalone is apparently not far off, and I guess the truly crazy gentleman at A1200.net is not far off with his keyboard mechanism. Assuming those both complete, here is what you have to do:

1. Put SD card containing AROS into Vampire standalone
2. Mount Vampire standalone in new A1200.net case
3. Route relevant cables to ports in back of case.
4. Attach keyboard mechanism, close case

And, if you are selling:

5. Slide A1200 and PSU into a couple pieces of bubble wrap
6. Insert into box.
7. Ship to customers.

As for the rest of your obstacles, they're not obstacles to shipping the above product. And you don't need that much money to do it.

[idrougge]

Originally Posted by wXR

As for the rest of your obstacles, they're not obstacles to shipping the above product. And you don't need that much money to do it.

But you need lots to buy it.

[wXR]

I do wonder how much it would be really. I suppose the big question is the standalone Vampire. One thing about matthey's multi-system vision is that it would probably drive hardware manufacture costs down considerably, since more could be produced.

[matthey]

Originally Posted by wXR

I don't really get your point about obstacles. A Vampire standalone is apparently not far off, and I guess the truly crazy gentleman at A1200.net is not far off with his keyboard mechanism. Assuming those both complete, here is what you have to do:

We already have small stand alone Amiga boards available today with the FPGA Arcade and Mist. Also, iComp's Reloaded Amiga 1200 boards will be available soon and maybe before the Vampire standalone boards. Some fit in a 1200 case better than others with the Reloaded being the best fit. Most of the features you requested are available or are planned to be available. The AmigaOS/AROS options out of the box are a big question mark and obstacle to Amiga sales. AmigaOS is a patch fest and license nightmare and AROS is much slower and less compatible.

Originally Posted by wXR

As for the rest of your obstacles, they're not obstacles to shipping the above product. And you don't need that much money to do it.

Actually, my listed obstacles are obstacles to shipping products. They greatly reduce the value of the products (less features and performance with more cost) which results in less products sold. There will be existing die hard Amiga users which buy these products and a few ex-Amiga users and high end retro gamers but we are talking about the result being low thousands of active 68k classic users still. This is not enough to develop software or an OS (open source or not) as can be seen with AmigaOS 4 and MorphOS. Neither of these camps have been successful at attracting outside users and most new development comes at a snails pace from a handful of sources (which are likely unprofitable as can be seen by Hyperion's rescue). There is no way to fix this problem without increasing the number of users and that starts with selling affordable mass produced products and marketing outside of the current Amiga community. No amount of software improvements would bring in new users without affordable hardware. The Amiga is in a death spiral as the user base declines. IMO, freeing the AmigaOS or greatly improving AROS does not fix the problem by itself.

Originally Posted by wXR

I do wonder how much it would be really. I suppose the big question is the standalone Vampire. One thing about matthey's multi-system vision is that it would probably drive hardware manufacture costs down considerably, since more could be produced.

I was considering an ASIC to cost between $500,000 and $1,000,000. Some people balk at this amount but it really isn't that much money anymore. Embedded partners could help defer the cost as the CPU could be sold by itself if it was 68k compatible enough and could have a long product life. There are very talented FPGA designers (including with CPU design experience) in the Amiga community which could help reduce engineering and design costs.

[Seiya]

the price must be competetive like Raspberry. Now many are very hit by Vampire with this price.
But inthe long run could be a problem.

[matthey]

Originally Posted by Seiya

the price must be competetive like Raspberry. Now many are very hit by Vampire with this price. But in the long run could be a problem.

The Raspberry Pi certainly set the bar for low priced hardware. The Raspberry Pi Foundation (a charitable foundation) was aiming for a low price to penetrate educational markets and the Pi was subsidized in many ways. I don't think it is necessary or practical to try competing in the $5-$50 U.S. range as this would require selling hundreds of thousands of a single chip SoC board. I think the $50-$150 range is a sweet spot which is obtainable (and affordable for most people) and could still sell well with the uniqueness and retro appeal of a product with the right features. I would look to sell at least 50,000 units.

[NovaCoder]

I'm interested to see where this goes, I actually think a stand-alone Apollo computer makes much more sense than attaching it to a 25 year old Amiga motherboard.

I expect that the new stand alone board (which is under development) will be more powerful than any of the Vampire boards.

It's also expect AROS 68k will get a lot more development because of these new boards along with an uplift of older 68k software and development tools (which has already been happening to a degree).

Interesting retro days

[wawa]

Originally Posted by NovaCoder

I actually think a stand-alone Apollo computer makes much more sense than attaching it to a 25 year old Amiga motherboard.

i tend to disagree. apart of the whole emotional load that such an old amiga device brings to the table, technically required or not, at least as justification of some legacy. there is a rational argument to this. and easy to prove just taking the historical development of natami/vampire/apollo initiative into account. i quote them all in the same context, where they consequently belong according to my view.

natami, which kicked off all that movement was too ambitious, too complete and too complex a device to start with. the softcore that came into play shortly thereafter needed to be tested against a stable reference. therefore a simple accelerator design was more appropriate to introduce and to debug, taking an opportunity of engagement, a crowd of early adopters happened to provide. this is where we are now. subsequently adding features to cpu implementation an introduction of a standalone is a logical next step. but then im still more attracted by a potential a1200 version, funny enough.

It's also expect AROS 68k will get a lot more development because of these new boards along with an uplift of older 68k software and development tools (which has already been happening to a degree).

i hope so. we have been patient enough. for nothing. still there are people who think, its less legit..

[matthey]

Originally Posted by NovaCoder

I'm interested to see where this goes, I actually think a stand-alone Apollo computer makes much more sense than attaching it to a 25 year old Amiga motherboard.

I expect that the new stand alone board (which is under development) will be more powerful than any of the Vampire boards.

The old hardware is cool but the new hardware can be so much more convenient and powerful.

Originally Posted by wawa

natami, which kicked off all that movement was too ambitious, too complete and too complex a device to start with. the softcore that came into play shortly thereafter needed to be tested against a stable reference. therefore a simple accelerator design was more appropriate to introduce and to debug, taking an opportunity of engagement, a crowd of early adopters happened to provide. this is where we are now. subsequently adding features to cpu implementation an introduction of a standalone is a logical next step. but then im still more attracted by a potential a1200 version, funny enough.

Some people said the 68000 CPU project was too ambitious at the time. It had 16 general purpose 32 bit registers when most CPUs had 2-6 GP 8 bit registers. It was so complex that it needed extensive micro-coding to get it done in a reasonable amount of time. It was a huge risk starting a new radically different architecture but when it came out it almost put Intel out of of business.

Some people said the Amiga was too ambitious before it was released. It was supposedly too complex and too costly. Steve Jobs supposedly said it was too much hardware. People didn't know what could be done with it. It was so far ahead of its time that people didn't know what to do with it. It almost failed because of the financing and very easily could have but it didn't.

I disagree about the Natami. It was ahead of its time just like the 68k and the Amiga. The hardware was designed by a professional. The direction was right IMO. There were obstacles though. Many of the same obstacles I mentioned. It could have been a success just like the 68k and Amiga could have failed with slightly different circumstances. It isn't about fate but about choice. Everyone make choices and it affects everyone around us.

[wawa]

i dont question thomas, his professionality. the obstacles were of logistical, not of technical nature.