Dreamcast Emulation on latest SOC's


vcoleiro1

Hardcore Member
Joined
Jan 23, 2011
Messages
4,685
As Arm SOC's are now getting pretty powerfull, are we at a stage where a Dreamcast emulator could be developed to run on one of the latest or soon to be released SOC's?


For example the Snapdragon S4 pro or the Exynos quad core or even the A15 based SOC's or soon to be released SOC's


I get the feeling these SOC's should give enough power?.
 
Last edited by a moderator:
Given the recent videos posted I'm going to say yes. Sonic Adventure was playing full speed most of the time on a pretty modest 1.2GHz dual core Cortex-A9 SoC. There are probably more demanding games but a 1.7+GHz Cortex-A15 is a lot more powerful. I doubt quad core helps you that much.
 
There are probably more demanding games but a 1.7+GHz Cortex-A15 is a lot more powerful. I doubt quad core helps you that much.
Which A15 SoCs will deliver good linux support and be available in low quantities?
 
Which A15 SoCs will deliver good linux support and be available in low quantities?

Who knows. The low quantity thing is often determined on a case by case basis so it's not something they'd advertise for everyone. These days wide applications are becoming a bigger thing with all the KS projects and what have you. Used to be nVidia wouldn't sell Tegra SoCs in batches below 100,000, that's obviously not the case anymore.
 
Exciting times ahead indeed. It would make for a killer EMU on any Pandora 2. But to be honest I give a P2 a very slim chance of eventuating.
 
As Arm SOC's are now getting pretty powerfull, are we at a stage where a Dreamcast emulator could be developed to run on one of the latest or soon to be released SOC's?


For example the Snapdragon S4 pro or the Exynos quad core or even the A15 based SOC's or soon to be released SOC's


I get the feeling these SOC's should give enough power?.
And we also need a good PS2 emulator too!!!
 
Exciting times ahead indeed. It would make for a killer EMU on any Pandora 2. But to be honest I give a P2 a very slim chance of eventuating.
Why is that?
ED and Craig seem to be at cross purposes.

Craig wants to sell android gameing devices with keyboards.

ED wants to make pocketable Linux PCs

Craig is secretive

ED is spends a lot of time keeping in touch with the community

Craig has a heap of refund requests

ED has pandoras shipping from stock.

I'm not saying they can't pull together to make a P2, but you can see why people are doubtful ;)
 
Last edited by a moderator:
Exciting times ahead indeed. It would make for a killer EMU on any Pandora 2. But to be honest I give a P2 a very slim chance of eventuating.
Why is that?
 
ED and Craig seem to be at cross purposes.
Craig wants to sell android gameing devices with keyboards.
ED wants to make pocketable Linux PCs
Craig is secretive
ED is spends a lot of time keeping in touch with the community
Craig has a heap of refund requests
ED has pandoras shipping from stock.
 
I'm not saying they can't pull together to make a P2, but you can see why people are doubtful ;)

Who says they have to pull together? Can't ED do it without Craig?
 
Ideally any Pandora successor should be a Craig and ED joint venture with each doing whatever plays to their strengths.


How the ICP2 release goes will greatly determine if Craig has learned any lessons from the Pandora fiasco - I for one am hoping that he has.
 
Ideally any Pandora successor should be a Craig and ED joint venture with each doing whatever plays to their strengths.


How the ICP2 release goes will greatly determine if Craig has learned any lessons from the Pandora fiasco - I for one am hoping that he has.
Not sure this is the ideal scenario. None of the current software guys (including Ed) want to work with Craig on another project. Craig is already breaking ICP2 kickstarter promises about documenting his progress online.


On the other hand, Craig still is a "vision guy" and I suspect that his Pandora2 vision (modular android console with top shelf controls) could actually be a winner.


With the kind of horsepower available to next gen SoC's, the era of needing highly optimized emulators, like the gp2x and pandora have, may be drawing to a close. The scores of android 'good enough' emulators will make consumers happy enough.


The torrents of android games will make the aging library of linux games seem puny and rusty and user-unfriendly by comparison. Craig is tired of being in a small niche, relying on a handful of mechanics to maintain a customized older model car.
 
Ideally any Pandora successor should be a Craig and ED joint venture with each doing whatever plays to their strengths.


How the ICP2 release goes will greatly determine if Craig has learned any lessons from the Pandora fiasco - I for one am hoping that he has.
Not sure this is the ideal scenario. None of the current software guys (including Ed) want to work with Craig on another project. Craig is already breaking ICP2 kickstarter promises about documenting his progress online.


On the other hand, Craig still is a "vision guy" and I suspect that his Pandora2 vision (modular android console with top shelf controls) could actually be a winner.


With the kind of horsepower available to next gen SoC's, the era of needing highly optimized emulators, like the gp2x and pandora have, may be drawing to a close. The scores of android 'good enough' emulators will make consumers happy enough.


The torrents of android games will make the aging library of linux games seem puny and rusty and user-unfriendly by comparison. Craig is tired of being in a small niche, relying on a handful of mechanics to maintain a customized older model car.
Surely there will always be a newer, more complex consoles to emulate?

There may be 'good enough' emus for old systems, but emulating recent(ish) systems will Always require large amounts of optimisation.

Who says that linux games are aging? Old games age. New games are new

Is there a huge onslaught of creative and productive tools flooding the android marketplace at 69p each? Scores of fully fledged IDEs available at the click of a button?
 
Last edited by a moderator:
I think if you polled people , and I believe we have already, the consensus would be a dual boot Android and Linux.  People want both.  You see the same discussion in other segments  like PC sticks. If you log onto those forums, you will see people are wanting dual boot Android and Linux on their PC Sticks. Seems it's the direction to go to.  It gives you the benefit of running Linux apps, games etc while also being able to play the latest Android more modern games and apps.  Win/Win

Having Android also covers your bases in case new emulators are developed soley for it. (Dreamcast etc)
 
With the kind of horsepower available to next gen SoC's, the era of needing highly optimized emulators, like the gp2x and pandora have, may be drawing to a close. The scores of android 'good enough' emulators will make consumers happy enough.
Not even remotely. An even higher level of optimization is needed for more powerful consoles. Going by the new video and the comments made by Craig he sees Pandora as an emulation machine more than anything else. The new one needs to emulate things the old one couldn't.
 
With the kind of horsepower available to next gen SoC's, the era of needing highly optimized emulators, like the gp2x and pandora have, may be drawing to a close. The scores of android 'good enough' emulators will make consumers happy enough.
Not even remotely. An even higher level of optimization is needed for more powerful consoles. Going by the new video and the comments made by Craig he sees Pandora as an emulation machine more than anything else. The new one needs to emulate things the old one couldn't.
Ok, I'll bite. Which consoles, specifically? Dreamcast is the only system that I have seen being discussed as coming into range with the A15, for example. Are there any emulators of more powerful consoles that you see running significantly better under P2 Linux than P2 Android? Will blitting to something like a DSS overlay still be an important performance factor?


My point was simply that a leap forward in CPU power will make a lot of currently glitchy Android emulators smoother, reducing the advantages that platform specific optimizations for Pandora would offer. Obvious, no?
 
What you said was that newer SoCs will be so powerful that there'll be no need to highly optimize emulators anymore. That has nothing to do with Android or Linux, nor Cortex-A15 in particular. PS2, Gamecube, Wii, etc emulators on PC are all highly optimized.

But yes, I do see there being systems that you'll want as much power as possible to emulate on a Cortex-A15. Not just Dreamcast, but DS, PSP, and Saturn for starters. In the future this will extend to newer consoles as well. And yes saving a few ms per frame on stuff like blits does matter.
 
^ So essentially the P2 will be able to run emulators for those 4 systems at fullspeed?


How many years into the future do you estimate before arm socs can realistically take a stab at PS2 or Game Cube emulation?
 
I'm not going to say anything decisive about what future hardware can and can't do, especially when we're speculating as to what that hardware's capabilities even are. What I'm saying is that for any emulators for those platforms any performance improvement will be helpful.

The only thing I know about PS2 or Gamecube emulation is that nothing announced from ARM slated for release before 2015 or so is going to be powerful enough to handle it well, and while I don't know much about what the next gen Atoms will do here I doubt it'll be fast enough either. Even ULV Haswells in tablets may struggle to be fast enough for those tasks. I think it'll be a long time because peak single threaded performance improvement in mobile CPUs is going to slow down now that the upper limits of acceptable power consumption are being hit, and improvements will be dictated more by manufacturing advancement.
 
Back
Top