Ninja

[OMars]

Well, last I heard was that samsung wouldn't provide a small amount of processors because they sell in bulk.

Oh and hi everybody, I'm baaaaaaaaaaack!

 

[nem]

Shouldn't you be learning how to code some super emulator by now?

 

[MWeston]

Hello. I have been following this board since sometime last year (around the time people started getting their first editions) and I finally decided to sign up tonight. I have put off buying a GP2X because I have a PSP with v1.50 and I just never felt the GP2X hardware was strong enough (not an invitation for flaming, just an opinion) to justify the extra cash at this point. Don't get me wrong, what people are doing with this system is absolutely phenominal!

I have also looked at that Ninja project and hoped something might come of it, but I have been toying with the idea of creating my own system since the Ninja project looks dead. I have the knowledge and access to the tools and manufacturing facilities to come up with a prototype, but the cost is another matter I think that is the main thing that stops anyone from going any further than a good idea. Once you look into all of the costs, it is pretty much impossible without hundreds of thousands of dollars at startup.

I have considered the idea of working on my own version of a replacement for the GP2X (not to say Gamepark won't do that themselves). By the time it would be finished, the 2X would be almost a couple years into its life cycle and something new would be welcome. I'd like to eventually get emulators at full speed for faster/more complex systems like PSX, Saturn, 3DO and maybe take a slight crack at the Dreamcast. That one might not work so well though!

I am fooling around/working on a design (in my spare time) with the Motorola i.MX31 with 128MB of DDR-SDRAM and a 480x320 16-bit LCD. I already have samples as Motorola is a great company to deal with in that respect. It sounds like a great project, but I wonder if anyone is interested in something like this or if the GP2X with some overclocking is all that is needed? The i.MX31 does have some moderate 3D capabilities (better than DS but not like a PSP) which might be a cool addition and an integrated ARM based FPU is always a welcome addition.

Maybe the Ninja is dead because it just isn't worth it. What do others think?

 

[sand_man]

I'd like to eventually get emulators at full speed for faster/more complex systems like PSX, Saturn, 3DO and maybe take a slight crack at the Dreamcast. That one might not work so well though!

Are you kidding?

 

[Alex.]

I'd like to eventually get emulators at full speed for faster/more complex systems like PSX, Saturn, 3DO and maybe take a slight crack at the Dreamcast. That one might not work so well though!

Are you kidding?

On his own handheld, not on the GP2X, Mr Negative. :lol:

- Alex

 

[WhiteFalcon]

I have considered the idea of working on my own version of a replacement for the GP2X (not to say Gamepark won't do that themselves). By the time it would be finished, the 2X would be almost a couple years into its life cycle and something new would be welcome. I'd like to eventually get emulators at full speed for faster/more complex systems like PSX, Saturn, 3DO and maybe take a slight crack at the Dreamcast. That one might not work so well though!

I am fooling around/working on a design (in my spare time) with the Motorola i.MX31 with 128MB of DDR-SDRAM and a 480x320 16-bit LCD. I already have samples as Motorola is a great company to deal with in that respect. It sounds like a great project, but I wonder if anyone is interested in something like this or if the GP2X with some overclocking is all that is needed? The i.MX31 does have some moderate 3D capabilities (better than DS but not like a PSP) which might be a cool addition and an integrated ARM based FPU is always a welcome addition.

Maybe the Ninja is dead because it just isn't worth it. What do others think?

Sounds great, stun us with pictures when it is ready

Good Luck!

 

[pantera6]

Hello. I have been following this board since sometime last year (around the time people started getting their first editions) and I finally decided to sign up tonight. I have put off buying a GP2X because I have a PSP with v1.50 and I just never felt the GP2X hardware was strong enough (not an invitation for flaming, just an opinion) to justify the extra cash at this point. Don't get me wrong, what people are doing with this system is absolutely phenominal!

I have also looked at that Ninja project and hoped something might come of it, but I have been toying with the idea of creating my own system since the Ninja project looks dead. I have the knowledge and access to the tools and manufacturing facilities to come up with a prototype, but the cost is another matter I think that is the main thing that stops anyone from going any further than a good idea. Once you look into all of the costs, it is pretty much impossible without hundreds of thousands of dollars at startup.

I have considered the idea of working on my own version of a replacement for the GP2X (not to say Gamepark won't do that themselves). By the time it would be finished, the 2X would be almost a couple years into its life cycle and something new would be welcome. I'd like to eventually get emulators at full speed for faster/more complex systems like PSX, Saturn, 3DO and maybe take a slight crack at the Dreamcast. That one might not work so well though!

I am fooling around/working on a design (in my spare time) with the Motorola i.MX31 with 128MB of DDR-SDRAM and a 480x320 16-bit LCD. I already have samples as Motorola is a great company to deal with in that respect. It sounds like a great project, but I wonder if anyone is interested in something like this or if the GP2X with some overclocking is all that is needed? The i.MX31 does have some moderate 3D capabilities (better than DS but not like a PSP) which might be a cool addition and an integrated ARM based FPU is always a welcome addition.

Maybe the Ninja is dead because it just isn't worth it. What do others think?

yes i hope you manage
but mayby we can hope for a dreamcast emu for the gp3x (anyone knowif it could be done)
ooo and mayby a n64 emulator, but we will have quite a few buttons

 

[MWeston]

[/quote]

Are you kidding?
[/quote]

As alex said, I definitely meant on my hardware or something next gen

I think 532MHz, integrated FPU and 0.8M triangles/second 3D performance would be perfect for all systems out there under the Dreamcast. Anything above that would be a challenge and just for pure fun!

Who knows? it amazes me that a 93MHz N64 can be emulated at 30fps on a 333MHz PSP, but I played it myself! It really breaks that 10:1 ratio, but the PSP is a quite a specifically built piece of hardware.

 

[Shikkers]

Are you kidding?

As alex said, I definitely meant on my hardware or something next gen

I think 532MHz, integrated FPU and 0.8M triangles/second 3D performance would be perfect for all systems out there under the Dreamcast. Anything above that would be a challenge and just for pure fun!

Who knows? it amazes me that a 93MHz N64 can be emulated at 30fps on a 333MHz PSP, but I played it myself! It really breaks that 10:1 ratio, but the PSP is a quite a specifically built piece of hardware.

The only reason it breaks the 10:1 ratio is because it has a similar processor to the N64 and a GPU...

 

[OMars]

Shouldn't you be learning how to code some super emulator by now?

Hey there, i really don't want to get mad so just stfu.........my good sir. B)

Hello. I have been following this board since sometime last year (around the time people started getting their first editions) and I finally decided to sign up tonight. I have put off buying a GP2X because I have a PSP with v1.50 and I just never felt the GP2X hardware was strong enough (not an invitation for flaming, just an opinion) to justify the extra cash at this point. Don't get me wrong, what people are doing with this system is absolutely phenominal!

I have also looked at that Ninja project and hoped something might come of it, but I have been toying with the idea of creating my own system since the Ninja project looks dead. I have the knowledge and access to the tools and manufacturing facilities to come up with a prototype, but the cost is another matter I think that is the main thing that stops anyone from going any further than a good idea. Once you look into all of the costs, it is pretty much impossible without hundreds of thousands of dollars at startup.

I have considered the idea of working on my own version of a replacement for the GP2X (not to say Gamepark won't do that themselves). By the time it would be finished, the 2X would be almost a couple years into its life cycle and something new would be welcome. I'd like to eventually get emulators at full speed for faster/more complex systems like PSX, Saturn, 3DO and maybe take a slight crack at the Dreamcast. That one might not work so well though!

I am fooling around/working on a design (in my spare time) with the Motorola i.MX31 with 128MB of DDR-SDRAM and a 480x320 16-bit LCD. I already have samples as Motorola is a great company to deal with in that respect. It sounds like a great project, but I wonder if anyone is interested in something like this or if the GP2X with some overclocking is all that is needed? The i.MX31 does have some moderate 3D capabilities (better than DS but not like a PSP) which might be a cool addition and an integrated ARM based FPU is always a welcome addition.

Maybe the Ninja is dead because it just isn't worth it. What do others think?

Seriously man, that is awesome. You could just try making prototypes, one at a time.

P.S. For you people who hate me, i am not pressuring anybody to do anything! Thank you

 

[Radek]

Hello. I have been following this board since sometime last year (around the time people started getting their first editions) and I finally decided to sign up tonight. I have put off buying a GP2X because I have a PSP with v1.50 and I just never felt the GP2X hardware was strong enough (not an invitation for flaming, just an opinion) to justify the extra cash at this point. Don't get me wrong, what people are doing with this system is absolutely phenominal!

I have also looked at that Ninja project and hoped something might come of it, but I have been toying with the idea of creating my own system since the Ninja project looks dead. I have the knowledge and access to the tools and manufacturing facilities to come up with a prototype, but the cost is another matter I think that is the main thing that stops anyone from going any further than a good idea. Once you look into all of the costs, it is pretty much impossible without hundreds of thousands of dollars at startup.

I have considered the idea of working on my own version of a replacement for the GP2X (not to say Gamepark won't do that themselves). By the time it would be finished, the 2X would be almost a couple years into its life cycle and something new would be welcome. I'd like to eventually get emulators at full speed for faster/more complex systems like PSX, Saturn, 3DO and maybe take a slight crack at the Dreamcast. That one might not work so well though!

I am fooling around/working on a design (in my spare time) with the Motorola i.MX31 with 128MB of DDR-SDRAM and a 480x320 16-bit LCD. I already have samples as Motorola is a great company to deal with in that respect. It sounds like a great project, but I wonder if anyone is interested in something like this or if the GP2X with some overclocking is all that is needed? The i.MX31 does have some moderate 3D capabilities (better than DS but not like a PSP) which might be a cool addition and an integrated ARM based FPU is always a welcome addition.

Maybe the Ninja is dead because it just isn't worth it. What do others think?

Ninja is a reasonable project. It couldn't get into a market because... you know you have to be competitive price wise with other systems and you will be having hard time doing so buying parts in a small quantity. Then there must be a case produced, all designed, tested, a firmware written...

Such small things alone when they all add up...

And how many units you would be able to sell initially as an unknow newcomer? 3, 5, 10?

Without an investor it's all a dead duck I'm afraid.

However...

There is a solution to those cost issues - adapt to use things what are already produced and common. So cheap even in a detail.
It's including an everything - for an example you could try to use existing cases of some handhelds.

Not ideal but possible and in a price reach.

I'd like to eventually get emulators at full speed for faster/more complex systems like PSX, Saturn, 3DO and maybe take a slight crack at the Dreamcast. That one might not work so well though!

Are you kidding?

The chip what he is using would be capable of such feats. It has over 500mips and some extra other capabilities as well. With decent programming why not? But it will not be that easy and lots of effort will be needed.

 

[OmarNawaz]

a nano-itx based console would be ideal

I promised myself I wouldnt start anything till I get my current projects finished (360Extender,P64,GP2X Un portablising)

Nano-its is so small like 12x12cm and pretty powerful low power usage low heat

www.mini-itx.com have nano-itx though they are kinda overpriced you can speak to other places for bulk prices and they are more competitive

really more like a custom mini laptop thingy with gaming controls

piccy ^ of the non header version 245gbp h34r:

 

[MWeston]

I have also considered the idea of including a large FPGA such as a Xilinx Spartan-3L XC3S1000L attaching it to the data bus of the processor and also tie it in with the inputs and LCD.

Imagine loading the core(s) for your favorite game system right into the FPGA and then pumping raw instructions directly to it. How cool is that?!!!


Yes, the costs on any project like this are very prohibitive but nothing is impossible with enough interest. I guess I am wondering if that interest exists.

 

[Radek]

I have also considered the idea of including a large FPGA such as a Xilinx Spartan-3L XC3S1000L attaching it to the data bus of the processor and also tie it in with the inputs and LCD.

Well... you are not only one with such ideas:
http://www.gp32x.de/board/index.php?showt...28042&st=30
http://www.gp32x.de/board/index.php?showt...432&hl=fpga

Then we have Squidge who is working on his FPGA type based project already.

Imagine loading the core(s) for your favorite game system right into the FPGA and then pumping raw instructions directly to it. How cool is that?!!!

Rather just only controller (dpad/joystick/buttons) inputs.

Yes, the costs on any project like this are very prohibitive but nothing is impossible with enough interest. I guess I am wondering if that interest exists.

Cost wise for a hardware? No as those midrange Spartan FPGAs are relatively cheap. The clock could be for such "hardwired" solution pretty low so with simple duallayer PCB board (cheap even for prototype type series) it should be possible. Then add some µC - there are some very dirt cheap Arm7 based ones to controll the fixed functionality.

I guess it might work and would be a perfect emulation machine.

But there is one and almost impossible to overcome issue - HDL code?
Nintendo will not give any documentation for its older machines. Sony too, the rest companies for 98 percent of probability as well.

Writing from scratch in HDL will be much harder than just going with faster CPU and making it in a C/assembler.
There are lots of "normal" source C code available already for emus.
There are ??? for VHDL/Verilog.

It'd be making more sense to have a hybrid machine - not strictly FPGA based but with some FPGA to handle some tasks. And then trying to implement in FPGA only those most demanding (and difficult for normal CPU) tasks.

But advances in mobile CPU might make it unnecessary as the CPU will be eventually able to handle everything without problems. At the other hand those PLD chips are improving as well (going cheaper, more dense and consuming less power) so...

Personally if I'd be doing such project like you are doing I'd chose FPGA + µC route. Because of that simple reason I can buy decent FPGAs at good prices even in detail and in packages what I could be able to solder on PCB myself. The I/O pins flexibility would simplify PCB design significantly too.

 

[OmarNawaz]

http://www.picocomputing.com/products/cards.php

this place has some fpgas was reading something about brute force cracking using a fpga on the cardbus to bypass OS using DRM

 

[Radek]

http://www.picocomputing.com/products/cards.php

this place has some fpgas was reading something about brute force cracking using a fpga on the cardbus to bypass OS using DRM

Those are very nice but Virtex isn't cheap. To be realist the Spartan series have to be used and they are quite powerfull anyway. For sure would outperform any Arm if a task could written in HDL.

 

[MWeston]

Well... you are not only one with such ideas:
http://www.gp32x.de/board/index.php?showt...28042&st=30
http://www.gp32x.de/board/index.php?showt...432&hl=fpga

Then we have Squidge who is working on his FPGA type based project already.

Hi Radek. I look forward to your responses. You are very insightful and offer intelligent responses. You have definitely put some time into thinking about these things that interest me as well.

I have seen those topics, but did not follow them to the end. Thanks for listing them here. I did a quick search for Squidge's FPGA project but didn't find any specific topic about it. I'm curious to learn more about it. If he has that design going already, I will focus on my fast ARM solution. In the end we may have two very creative and interesting designs to play with!

Cost wise for a hardware? No as those midrange Spartan FPGAs are relatively cheap. The clock could be for such "hardwired" solution pretty low so with simple duallayer PCB board (cheap even for prototype type series) it should be possible. Then add some µC - there are some very dirt cheap Arm7 based ones to controll the fixed functionality.

I guess it might work and would be a perfect emulation machine.

But there is one and almost impossible to overcome issue - HDL code?
Nintendo will not give any documentation for its older machines. Sony too, the rest companies for 98 percent of probability as well.

Writing from scratch in HDL will be much harder than just going with faster CPU and making it in a C/assembler.
There are lots of "normal" source C code available already for emus.
There are ??? for VHDL/Verilog.

Oh, I agree with you. You won't find any datasheet handouts from the big boys, but the truth is that most, if not all, of these systems have emulators written for them already. One can take the C code and convert it into VHDL through some crafty reverse engineering. No one could say it would be easy and much of it would be done from scratch, although, I seem to recall HDL code for cores like the Z80 and 6502 are already out there?

There is also that fact that a 500MHz processor could probably make any low to midrange system run at 100% and so why even bother with the FPGA? I guess the only answer is: for fun or just to be different.

It'd be making more sense to have a hybrid machine - not strictly FPGA based but with some FPGA to handle some tasks. And then trying to implement in FPGA only those most demanding (and difficult for normal CPU) tasks.

But advances in mobile CPU might make it unnecessary as the CPU will be eventually able to handle everything without problems. At the other hand those PLD chips are improving as well (going cheaper, more dense and consuming less power) so...

Personally if I'd be doing such project like you are doing I'd chose FPGA + µC route. Because of that simple reason I can buy decent FPGAs at good prices even in detail and in packages what I could be able to solder on PCB myself. The I/O pins flexibility would simplify PCB design significantly too.

Yes, originally that is what I meant was to combine the FPGA with the fast ARM processor. Then you have the speed to emulate the "easier" systems in C and the hardware to write optimized algorithms for the bigger systems. Then of course, as I mentioned before, you could also re-write any system to run in the FPGA for fun. The one thing that gets me at this point is not the cost (ah, who cares, a prototype will be expensive anyway, what's a few more parts! ). It's the power consumption! I would go with a 4AA design, but even still, it could be a real power hog.

Maybe next gen SOC's will take a lesson from Sony and include some programmable logic in with all of the other new features that are finally getting in (like true High Speed USB 2.0 and DDR2, woohoo!).

 

[OMars]

Well... you are not only one with such ideas:
http://www.gp32x.de/board/index.php?showt...28042&st=30
http://www.gp32x.de/board/index.php?showt...432&hl=fpga

Then we have Squidge who is working on his FPGA type based project already.

Hi Radek. I look forward to your responses. You are very insightful and offer intelligent responses. You have definitely put some time into thinking about these things that interest me as well.

I have seen those topics, but did not follow them to the end. Thanks for listing them here. I did a quick search for Squidge's FPGA project but didn't find any specific topic about it. I'm curious to learn more about it. If he has that design going already, I will focus on my fast ARM solution. In the end we may have two very creative and interesting designs to play with!

Cost wise for a hardware? No as those midrange Spartan FPGAs are relatively cheap. The clock could be for such "hardwired" solution pretty low so with simple duallayer PCB board (cheap even for prototype type series) it should be possible. Then add some µC - there are some very dirt cheap Arm7 based ones to controll the fixed functionality.

I guess it might work and would be a perfect emulation machine.

But there is one and almost impossible to overcome issue - HDL code?
Nintendo will not give any documentation for its older machines. Sony too, the rest companies for 98 percent of probability as well.

Writing from scratch in HDL will be much harder than just going with faster CPU and making it in a C/assembler.
There are lots of "normal" source C code available already for emus.
There are ??? for VHDL/Verilog.

Oh, I agree with you. You won't find any datasheet handouts from the big boys, but the truth is that most, if not all, of these systems have emulators written for them already. One can take the C code and convert it into VHDL through some crafty reverse engineering. No one could say it would be easy and much of it would be done from scratch, although, I seem to recall HDL code for cores like the Z80 and 6502 are already out there?

There is also that fact that a 500MHz processor could probably make any low to midrange system run at 100% and so why even bother with the FPGA? I guess the only answer is: for fun or just to be different.

It'd be making more sense to have a hybrid machine - not strictly FPGA based but with some FPGA to handle some tasks. And then trying to implement in FPGA only those most demanding (and difficult for normal CPU) tasks.

But advances in mobile CPU might make it unnecessary as the CPU will be eventually able to handle everything without problems. At the other hand those PLD chips are improving as well (going cheaper, more dense and consuming less power) so...

Personally if I'd be doing such project like you are doing I'd chose FPGA + µC route. Because of that simple reason I can buy decent FPGAs at good prices even in detail and in packages what I could be able to solder on PCB myself. The I/O pins flexibility would simplify PCB design significantly too.

Yes, originally that is what I meant was to combine the FPGA with the fast ARM processor. Then you have the speed to emulate the "easier" systems in C and the hardware to write optimized algorithms for the bigger systems. Then of course, as I mentioned before, you could also re-write any system to run in the FPGA for fun. The one thing that gets me at this point is not the cost (ah, who cares, a prototype will be expensive anyway, what's a few more parts! ). It's the power consumption! I would go with a 4AA design, but even still, it could be a real power hog.

Maybe next gen SOC's will take a lesson from Sony and include some programmable logic in with all of the other new features that are finally getting in (like true High Speed USB 2.0 and DDR2, woohoo!).

I have read your posts, but it is still obscure to me whether you are going to try to make this thing or not. If the answer is yes i have a sketchbook full of console designs, if you want me to scan them for you...... :huh: Ok then...bye for now.

 

[Radek]

Well... you are not only one with such ideas:
http://www.gp32x.de/board/index.php?showt...28042&st=30
http://www.gp32x.de/board/index.php?showt...432&hl=fpga

Then we have Squidge who is working on his FPGA type based project already.

Hi Radek. I look forward to your responses. You are very insightful and offer intelligent responses. You have definitely put some time into thinking about these things that interest me as well.

You know probably about my little project of stick replacement, don't you?
Simple as it's looking but there were many considerations. It ended using an µC chip and some parts from some cell phone.
(and it taken lots of time so find suitable µC because of needed voltage tolerance, power consumption and cost of getting a programmer for it plus time needed to learning to program it)

All that because of parts availability and size constrains.

To be really able to make something what could have a chance of success (or even a showing a prototype) there is no place for a dreaming. No, no that dreaming is bad but to be able materialize something anyone has to use things what are realistic to get.

Look at the Ninja - Samsung wasn't willing to provide chips in smaller quantity. Is Samsung bad or there was a design mistake to use a chip what couldn't be bought in such limited numbers?

You decide.

Imho to get results should be a major priority. If Samsung couldn't provide a needed element then it's time to adapt to somethng else.

I can get some 70MHz Arm7 chips and Xilinx Spartant XC3S500E rather easily and CHEAPLY. I can make (via some other company) dual sided PCBs plus soldering these chips at acceptable price even for single digit unit production.

So I could afford to make prototypes and low scale production.
(at least I could afford to take risks to try with such project as it'd be within my capabilities)

I have seen those topics, but did not follow them to the end. Thanks for listing them here. I did a quick search for Squidge's FPGA project but didn't find any specific topic about it. I'm curious to learn more about it. If he has that design going already, I will focus on my fast ARM solution. In the end we may have two very creative and interesting designs to play with!

Well... Squidge is more interesting in making his own machine than an emulating some other. Anyway try ask him him for details.

And don't be afraid of FPGA or even much simpler CPLD chips.
They can simplify a design drastically!
(even used only for rerouting signals - you DON'T want to resort to multilayer PCB as the costs will kill you unless you have lots of money to spend). So even 2$ CPLD chip can be very handy.

Cost wise for a hardware? No as those midrange Spartan FPGAs are relatively cheap. The clock could be for such "hardwired" solution pretty low so with simple duallayer PCB board (cheap even for prototype type series) it should be possible. Then add some µC - there are some very dirt cheap Arm7 based ones to controll the fixed functionality.

I guess it might work and would be a perfect emulation machine.

But there is one and almost impossible to overcome issue - HDL code?
Nintendo will not give any documentation for its older machines. Sony too, the rest companies for 98 percent of probability as well.

Writing from scratch in HDL will be much harder than just going with faster CPU and making it in a C/assembler.
There are lots of "normal" source C code available already for emus.
There are ??? for VHDL/Verilog.

Oh, I agree with you. You won't find any datasheet handouts from the big boys, but the truth is that most, if not all, of these systems have emulators written for them already. One can take the C code and convert it into VHDL through some crafty reverse engineering. No one could say it would be easy and much of it would be done from scratch, although, I seem to recall HDL code for cores like the Z80 and 6502 are already out there?

Such 8-bits CPUs are easy enough to emulate using Arm. Don't waste your time doing them using FPGA. FPGA should be used rather for custom chipsets emulation.
(except there enough resources left in FPGA)

There is also that fact that a 500MHz processor could probably make any low to midrange system run at 100% and so why even bother with the FPGA? I guess the only answer is: for fun or just to be different.

If you can buy 500MHz units easily then go for them!! But consider all variables including PCB complexity - FPGA/CPLD are flexible in pins I/O.
As I said before it's nothing wrong to put a CPLD to just reroute some signals so you might not have to resort to multilayer PCB.

It'd be making more sense to have a hybrid machine - not strictly FPGA based but with some FPGA to handle some tasks. And then trying to implement in FPGA only those most demanding (and difficult for normal CPU) tasks.

But advances in mobile CPU might make it unnecessary as the CPU will be eventually able to handle everything without problems. At the other hand those PLD chips are improving as well (going cheaper, more dense and consuming less power) so...

Personally if I'd be doing such project like you are doing I'd chose FPGA + µC route. Because of that simple reason I can buy decent FPGAs at good prices even in detail and in packages what I could be able to solder on PCB myself. The I/O pins flexibility would simplify PCB design significantly too.

Yes, originally that is what I meant was to combine the FPGA with the fast ARM processor. Then you have the speed to emulate the "easier" systems in C and the hardware to write optimized algorithms for the bigger systems. Then of course, as I mentioned before, you could also re-write any system to run in the FPGA for fun. The one thing that gets me at this point is not the cost (ah, who cares, a prototype will be expensive anyway, what's a few more parts! ). It's the power consumption! I would go with a 4AA design, but even still, it could be a real power hog.

No if it's FPGA/CPLD would work at low clockrates. These are basically like a hardwired chips and they have perfomance advantage over 10times (if not more) againsts the same task performed via normal CPU at the same clockrate. If balanced well it shouldn't be power hog.

Maybe next gen SOC's will take a lesson from Sony and include some programmable logic in with all of the other new features that are finally getting in (like true High Speed USB 2.0 and DDR2, woohoo!).

Better think using parts what are available and not in big numbers for an individual consumers.
(and the ones you could make a PCB for)

I have an idea what looks realistic enough but and the word is a "hybrid".
Even better if... (I wrote about a core of my idea already somewhere)

 

[OMars]

I can get some 70MHz Arm7 chips and Xilinx Spartant XC3S500E rather easily and CHEAPLY. I can make (via some other company) dual sided PCBs plus soldering these chips at acceptable price even for single digit unit production.

So I could afford to make prototypes and low scale production.
(at least I could afford to take risks to try with such project as it'd be within my capabilities)

That's cool, even 70mhz seems like enough, because as long as developers are willing to program for it, power is not that essential...case and point: the gameboy advance - 16mhz and (i believe 256kb ram :huh: ), yet it has a lot of great emus. You seem to be good with this hardware engineering stuff. I think(again, i'm not forcing anybody)you should try to make this work. Good luck if you decide to though!