Theoretically Speaking, Is The Pandora At Least Capable Of Dreamcast E

[TaG]

see here for plan from drkIIRaziel, author.
So he is going to release it after a rewrite targetted for embedded devices? I hope it doesn't take too much time to do, I'm a bit impacient

Ok, the full post, April 18th 2008, was:QUOTE
There are varius complications around the source code so opensourcing isnt that simple.I do plan to release source, but that requires massive amount of work to be possible.nullDC is designed and implemented around x86/Windows and relies havily on features like managed memory, file mapping and other OS services that are typicaly not avaiable on consoles.I plan to start a rewrite from scratch that will target embeded enviroments (psp lite for now) that will be opensource
so I assume the psp lite version will be opensource in time. It is still considered alpha.

 

[quadomatic]

see here for plan from drkIIRaziel, author.
So he is going to release it after a rewrite targetted for embedded devices? I hope it doesn't take too much time to do, I'm a bit impacient

Ok, the full post, April 18th 2008, was:QUOTE
There are varius complications around the source code so opensourcing isnt that simple.I do plan to release source, but that requires massive amount of work to be possible.nullDC is designed and implemented around x86/Windows and relies havily on features like managed memory, file mapping and other OS services that are typicaly not avaiable on consoles.I plan to start a rewrite from scratch that will target embeded enviroments (psp lite for now) that will be opensource
so I assume the psp lite version will be opensource in time. It is still considered alpha.


So I imagine it will be a while before the source is released then. Maybe someone can turn him onto Pandora? There's a decent chance he'd be interested I suppose. We could at least send him the specs of the Pandora and ask what he thinks is possible.

I am still airing on the side of dreamcast emulation being a pipe dream, but since he's actually bothering to try DC emulation on PSP, maybe he'd be willing to give Pandora a shot.

 

[Chip]

Very impressive videos. That the author was able to boot the DC BIOS and have the system menus working on something as limited as the PSP is no small feat. I am genuinely impressed at his prowess. Those videos show that he has at least removed the various windows dependencies from the original NullDC emulator, which is nothing to sneeze at. If he follows through with his stated intention to release the source, then a Panodra port is very possible.

A playable port, however, is another thing entirely. Before anybody soils themselves in anticipation of Crazy Taxi or Jet Set Radio on the Pandora, please try to keep this in perspective. All these videos show is the DC system menu running at single digit frame rates. There may well be additional improvements that could be made to the speed of emulation, but I wouldn't hold my breath for improvements so drastic as to make any commercial game playable on the PSP. Even the Pandora, with its significantly faster CPU and extra RAM would still crawl with the code as it now stands (assuming it could even be ported and what we're seeing isn't heavily optimized for the PSP specifically).

It's an interesting proof of concept, but that's all it is at the moment. If the source is released it will open the door for devs to get something working on Pandora hardware, but the chance of playable speeds is still pretty slim.

 

[quadomatic]

Very impressive videos. That the author was able to boot the DC BIOS and have the system menus working on something as limited as the PSP is no small feat. I am genuinely impressed at his prowess. Those videos show that he has at least removed the various windows dependencies from the original NullDC emulator, which is nothing to sneeze at. If he follows through with his stated intention to release the source, then a Panodra port is very possible.

A playable port, however, is another thing entirely. Before anybody soils themselves in anticipation of Crazy Taxi or Jet Set Radio on the Pandora, please try to keep this in perspective. All these videos show is the DC system menu running at single digit frame rates. There may well be additional improvements that could be made to the speed of emulation, but I wouldn't hold my breath for improvements so drastic as to make any commercial game playable on the PSP. Even the Pandora, with its significantly faster CPU and extra RAM would still crawl with the code as it now stands (assuming it could even be ported and what we're seeing isn't heavily optimized for the PSP specifically).

It's an interesting proof of concept, but that's all it is at the moment. If the source is released it will open the door for devs to get something working on Pandora hardware, but the chance of playable speeds is still pretty slim.

I agree with you, but I'm hoping Raziel can prove us wrong. Pandora does run circles around the PSP (right?). I know its wishing for A LOT, but Dreamcast emulation would make me shit bricks.

Maybe craigix could get in contact with him and offer to put him on the list for a dev board?

Sigh...all this talk about DC makes me want to plug my Dreamcast back in and play some Jet Grind Radio...or actually try to finish Shenmue. The idea of finishing Shenmue 1 and 2 always makes me sad though, since there aren't any more Shenmue games to play in the near future.

 

[TaG]

A playable port, however, is another thing entirely. Before anybody soils themselves in anticipation of Crazy Taxi or Jet Set Radio on the Pandora, please try to keep this in perspective. All these videos show is the DC system menu running at single digit frame rates.
Awww Chip. Couldn't you have let us dream just a little longer?

Still we can hope. After all, thats what is left in the box. Fitting really.

 

[nubie]

Whenever the subject of what systems can be emulated on the Pandora comes up, the Dreamcast always seems to get filed into the same "not gonna happen" category as the PS2/Xbox/Gamecube.

Now I don't claim to be the most knowledgeable guy out there when it comes to hardware, but from what I understand the Dreamcast wasn't really that close to the PS2/GC/Xbox in terms of processing power, and its hardware wasn't overly complex either, making it much less difficult to emulate.

Moreover, the fact that the Pandora contains a more advanced version of the PowerVR chip that the Dreamcast uses, should make emulating it less of an obstacle. So most of the difficulty would come from emulating the SH-4, but with a fast recompiler and the Pandora overclocked to 900 MHz I don't think this is outside the realm of possibility; the existence of Chankast (runs many games full speed with only a 1.5 GHz Pentium 4) should at least prove that getting some games to run at decent speeds on the Pandora shouldn't be a total pipe dream.

I understand that a decent emulator would probably have to be written entirely from scratch and that's probably the main reason that it isn't likely to happen, but do any developers think that the Pandora is technically capable of decent Dreamcast emulation?

Paragraphs exist for a reason.

"the existence of Chankast (runs many games full speed with only a 1.5 GHz Pentium 4) should at least prove that getting some games to run at decent speeds on the Pandora shouldn't be a total pipe dream. "

This is entirely flawed logic, a Pentium 4 has no relevance here.

As far as the GC/PS2/Xbox debate goes: the DC had more usable graphical power than the PS2 (the PS2 was marketed in an extreme manner, and did have more theoretical raw processing power), look it up. Thus more games are likely to be using more of the processing power than a comparable PS2 game (early released games for PS2).

As for the GC and Xbox, yes they are more powerful, but in the large scheme of things the DC is much closer to them than you might guess.

I would love to see a DC emu since the GPU's are similar, but I think that they are more dissimilar than is commonly believed.

If someone was to write a dynamic compiler for the OMAP and give it a good shot there is a small possibility, but I wouldn't expect it to be any better than psx emulation on the GP2X, a novelty at best due to lack of resources.

 

[Kayday]

Here's the link to the thread talking about the PSP port :

http://forums.ngemu.com/nulldc-discussion/...nulldc-psp.html

Someone did mention the Pandora in the thread so maybe Raziel has heard of the Pandora. Hopefully craigix can get in touch with him and see if he's interested.

 

[SpacemanSpiff]

As far as the GC/PS2/Xbox debate goes: the DC had more usable graphical power than the PS2 (the PS2 was marketed in an extreme manner, and did have more theoretical raw processing power), look it up. Thus more games are likely to be using more of the processing power than a comparable PS2 game (early released games for PS2).

How do you figure that? If you look at real world polygon output, very few DC games even approached 1 million polygons per second, while even early PS2 software dwarfed that number. (Source: vertices per second reported when running NullDC, polygons per frame reported by PCSX2) I don't know what other benchmarks you could be using to make that claim, but its pretty clear to me that the PS2 hardware was capable of doing things that the DC couldn't even dream of.

 

[CyruzDraxs]

Technically speaking, DC on the Pandora IS doable. However; it would not be a simple task at all.

There is many hardware similarities that would allow many instructions to be handled un-emulated and just put through a controlled environment. Though this COULD be done, it would require many, many hours of development by a very, very knowledgeable developer as current open source alternatives are just too different to be used for non-emulated instruction controlling. If some Godly developer came along tomorrow and picked up the project we still probably wouldn't see any playable progress from it for at least a year.

I'll just put it this way; it's safe to say that we may see open source games of equivalent graphical levels as the DC ported to the Pandora, but I can say with almost complete certainty that we will never see a DC emulator.

 

[YenningComity]

If someone really wants to know if pandora can emulate a DC then they should look at the instruction sets for the PowerVR2 and SGX *aka PowerVR5*. If they are like Intel, it could be possible to not emulate the code and use the gpu in pandora. If not they don't have the same calls, don't expect full speed.

Even if this would work you would still have a 600mhz ARM cpu doing the task of the rest of the console, no small feat. If someone were really savy they could offload a little work to the DSP, but this would take a lot of time.

My question, is even if there were a few programmers with the skill and time, would we want them to waste it on DC emulation? I am a big fan of some of the games released on the console, but it seems like a bad idea. Theoretically speaking you could do Naomi *might be a few mb short of ram for some games*, and atomis wave, but that's it. If Pandora fails, much of the code is useless, if there is a successor to pandora that doesn't use the next PowerVR chip, it is useless. You can't do Naomi 2 games because it had twin gpu's, and 4x the ram of the dreamcast. You cannot use any games that required a gd rom because sd cards won't read fast enough to run. All this of course ignoring the fact you must break the encryption which won't happen because the Mame team is probably the only ones with the resources and knowledge to do, and they won't because its still a money maker.

Sorry to be so negative, but unless someone just wants to do it to learn, it won't happen. If someone ports nulldc to see I won't be surprised, but the dedication it would take to get it playable isn't likely.

 

[quadomatic]

If someone really wants to know if pandora can emulate a DC then they should look at the instruction sets for the PowerVR2 and SGX *aka PowerVR5*. If they are like Intel, it could be possible to not emulate the code and use the gpu in pandora. If not they don't have the same calls, don't expect full speed.

Even if this would work you would still have a 600mhz ARM cpu doing the task of the rest of the console, no small feat. If someone were really savy they could offload a little work to the DSP, but this would take a lot of time.

My question, is even if there were a few programmers with the skill and time, would we want them to waste it on DC emulation? I am a big fan of some of the games released on the console, but it seems like a bad idea. Theoretically speaking you could do Naomi *might be a few mb short of ram for some games*, and atomis wave, but that's it. If Pandora fails, much of the code is useless, if there is a successor to pandora that doesn't use the next PowerVR chip, it is useless. You can't do Naomi 2 games because it had twin gpu's, and 4x the ram of the dreamcast. You cannot use any games that required a gd rom because sd cards won't read fast enough to run. All this of course ignoring the fact you must break the encryption which won't happen because the Mame team is probably the only ones with the resources and knowledge to do, and they won't because its still a money maker.

Sorry to be so negative, but unless someone just wants to do it to learn, it won't happen. If someone ports nulldc to see I won't be surprised, but the dedication it would take to get it playable isn't likely.

Oh well, if we can't do NAOMI...then the only thing we can emulate is:

To the next generation! Portable Virtua Fighter 5 FTW!.... :unsure:

 

[icurafu]

Hopefully because of the hard work on nullDC to get it running on embedded devices as opensource, we'll see mobile DC on a future homebrew system. (Such as the Pandora 2, GP4X or the Nintendo EndGame)

 

[quadomatic]

Hopefully because of the hard work on nullDC to get it running on embedded devices as opensource, we'll see mobile DC on a future homebrew system. (Such as the Pandora 2, GP4X or the Nintendo EndGame)

wtf...Nintendo EndGame?

 

[AireTamStorm]

Hopefully because of the hard work on nullDC to get it running on embedded devices as opensource, we'll see mobile DC on a future homebrew system. (Such as the Pandora 2, GP4X or the Nintendo EndGame)

Or the Sony PlayStation Rolly. Offload some of the work to the motors....

 

[icurafu]

Oh Rolly? o0

 

[realyst]

I prefer the Nintendo SharkJumper or the Microsoft PleaseWork.

On topic, however, I'd like to see what this guy could squeeze out of the Panda. Even if it's isn't playable.

 

[YenningComity]

It seems like I should correct myself http://www.arcade-renaissance.com/2008/02/...-striker-2.html seems Naomi has already been started.

I am wondering if model 1 and 2 emulation would be possible. Model 3 was more advanced then a DC, but the original DOA and Virtua Fighters in arcade quality would be nice.

 

[quadomatic]

It seems like I should correct myself http://www.arcade-renaissance.com/2008/02/...-striker-2.html seems Naomi has already been started.

I am wondering if model 1 and 2 emulation would be possible. Model 3 was more advanced then a DC, but the original DOA and Virtua Fighters in arcade quality would be nice.



I think for now we're better off just keeping our fingers crossed that a good developer decides to give dc emulation on the Pandora a shot.

 

[atomicthumbs]

Theoretically the Pandora will be powerful to emulate itself at full speed, thus producing more Pandoras which will take away the disadvantage of the initial production run of only 3000 units

 

[quadomatic]

Theoretically the Pandora will be powerful to emulate itself at full speed, thus producing more Pandoras which will take away the disadvantage of the initial production run of only 3000 units

NOoOoOoO

You've broken the space time continuum! We will all cease to exist!!!!