Pandora 2, A Year Or So From Now?

[s33k3rgr]

I would be happy if the next openpandora runs the Dreamcast emulator perfectly. I would be even happier if the current openpandora runs that emulator ok. Am i asking too much?

 

[zlice0]

I only read the first few post, but I would totally buy a new Pandora with the cortex-9. Dual 1GHz, most certainly more memory and the capabilities on the cortex9 are better. Why let all the big names have the fun? Pandora would just be left in the dust if everyone gets a new OMAP4 next year. Open source is the way to go (opinion). And with people who now know what they are doing and where the pit falls are, a second Pandora would be better hardware and software released in a better time frame. I've been wanting to get into embedded for a while too, I'd be on the bandwagon and help out if I could.

 

[spud42]

That being said, a few things I would like to see in a future Pandora successor:

An ARM powerful enough to emulate 6th Gen home consoles while maintaining code compatibility with Cortex A8 binaries.
100GB+ of internal storage and Dual SDXC slots(All the extra space will be needed to make DVD-based games viable).
Clamshell to Tablet transformation(Screen flips out 180 degrees and then slides down over the controls).
A built-in camera would be nice, but ONLY if it was comparable to dedicated digital cameras/camcorders.
Line-in with recording capability for both audio and video sources.

I'd like to see it make ramen, too, but that's not practical.
You probably won't see 100 GB of storage unless someone, right now, is discovering a way to make flash vastly cheaper or hard drives vastly smaller, cheaper, more power-efficient, and more durable. Edit: Actually, in 5 years, with Moore's law... This might happen. But that's quite far out to be thinking.
Tablets are apparently patented. What a dick move, right? Right. Dicks.
The camera would be nice since apparently cell phones have fantastic cameras, but I couldn't fucking download the video from my last one because it was a piece of shit.
I haven't even seen desktop PCs with video-in recorders. They're a really cool feature, but most people don't need them, so they're considered peripheral. I have no clue how a handheld with a cheap ARM chipset would have one built-in.

his original wish list had CF slot on it... there are already 100Gb CF cards available.i remember talking about it on a ipaq Hx4700 forum about a year ago. it was for high end digital cameras but was CF. The CF1 standard allows addressing of up to 137Gb . found it
i agree the price is steep!

 

[WizardStan]

But but but but... in the space used by 2 SD slots, you could have 6 micro-SD slots!

And for the cost of those 6 micro-SD cards, you could get 12 full sized SD cards and swap them around when needed.

 

[hch]

NO NO NO NO GOOD GOD NO! Micro SD is great for devices that don't have the space for a full sized SD card. If the device can fit a full sized SD card, there is absolutely no reason to put in a micro and many reasons not to.

But but but but... in the space used by 2 SD slots, you could have 6 micro-SD slots!

yay, and have a complete failover raid-5 with hotspare. right in your pocket.

 

[Exophase]

The OMAP3 (and likely OMAP4) only has the capability to control two SD slots.

OMAP3 has three SD/MMC interfaces. Pandora uses one of them for WiFi over SDIO. It might be possible to interface with SPI instead, which is basically the same, and OMAP3 has 4 SPI controllers - not sure how they're being used on Pandora. OMAP4 is largely unknown right now.

OMAP4 doesn't seem like big enough of a deal for pandora 2 to be released on that platform, it's such a miniscule upgrade (in comparison to other consoles generations') that it's sort of a waste, it's not even a doubling of performance. If other consoles upgraded like this, we would be up to playstation 6.

Pandora is a console, you don't upgrade your console every single year for something bigger and better that's available. Maybe 2 years from now, OMAP5 might be a big enough upgrade for it to matter.

But it isn't actually a console in most senses. A console is driven by software sales and therefore it's in the company's best interests to keep the hardware on the market for as long as possible. This is especially true when backwards compatibility is difficult, and it often is.

craigix has stated before that he wants to do something like what Apple is doing with iPhone, which means hardware revisions spaced around ~2 years. Even DS and PSP have been doing a bit of this, or more than non-handheld consoles have. Home console release cycles are closer to 5 years.

I don't know how you determined that OMAP4 is "not even a doubling of performance" though. It depends on what you're looking at, but at least in terms of CPU I would rank 2 1GHz Cortex-A9 cores as well beyond double what 1 600MHz Cortex-A8 can do. There almost definitely won't be a viable OMAP5 option in 2 years.

You guys talk a lot about "drop in upgrades", but something around OMAP4 would probably be a substantial board redesign.

 

[Ravnos]

The OMAP3 (and likely OMAP4) only has the capability to control two SD slots.

OMAP3 has three SD/MMC interfaces. Pandora uses one of them for WiFi over SDIO. It might be possible to interface with SPI instead, which is basically the same, and OMAP3 has 4 SPI controllers - not sure how they're being used on Pandora. OMAP4 is largely unknown right now.

OMAP4 doesn't seem like big enough of a deal for pandora 2 to be released on that platform, it's such a miniscule upgrade (in comparison to other consoles generations') that it's sort of a waste, it's not even a doubling of performance. If other consoles upgraded like this, we would be up to playstation 6.

Pandora is a console, you don't upgrade your console every single year for something bigger and better that's available. Maybe 2 years from now, OMAP5 might be a big enough upgrade for it to matter.

But it isn't actually a console in most senses. A console is driven by software sales and therefore it's in the company's best interests to keep the hardware on the market for as long as possible. This is especially true when backwards compatibility is difficult, and it often is.

craigix has stated before that he wants to do something like what Apple is doing with iPhone, which means hardware revisions spaced around ~2 years. Even DS and PSP have been doing a bit of this, or more than non-handheld consoles have. Home console release cycles are closer to 5 years.

I don't know how you determined that OMAP4 is "not even a doubling of performance" though. It depends on what you're looking at, but at least in terms of CPU I would rank 2 1GHz Cortex-A9 cores as well beyond double what 1 600MHz Cortex-A8 can do. There almost definitely won't be a viable OMAP5 option in 2 years.

You guys talk a lot about "drop in upgrades", but something around OMAP4 would probably be a substantial board redesign.

I'm guessing when they say "drop in upgrades", they mean that the board is the only component needing that redesign. It would be substantial, but at the same time, there would be no need to redesign the case, change to new nubs, etc. All the little stuff that caused big delays. Unless the next Pandora is also going to be a phone (which would be sweet if it was done right), the case and everything else is pretty much perfect for what it is.

But maybe I'm reading too much into it, maybe they think you can just change the SOC and then the magic happens.

 

[Exophase]

I really hope it's not turned into a phone. That'd over $100 to the cost for something a lot of people don't want. Seems better to use an add-on for that.

It's been a bit mixed, as to whether people are talking about the whole board or just the CPU. It's a given that the case and components will be the same, since craigix has expressed this.

The other usage of "drop in" has been the suggestion that Open Pandora sells a motherboard upgrade for the Pandora, which I think would not be a good idea, although I suppose it'd be very attractive to some.

 

[Awakening]

Skipping all previous posts...

I really wonder how many existing owners would pay for another Pandora just to get a power upgrade. Those who buy a Pandora just for emulation will be perfectly happy with what they got and will play until it breaks. A power upgrade will not add a whole lot more, it may attract new users but developers would have to deal with 2 different hardwares.

 

[SNESFAN]

Skipping all previous posts...

I really wonder how many existing owners would pay for another Pandora just to get a power upgrade. Those who buy a Pandora just for emulation will be perfectly happy with what they got and will play until it breaks. A power upgrade will not add a whole lot more, it may attract new users but developers would have to deal with 2 different hardwares.

i dump roughly 2-3k a year just on gadgets that don't really do that much different stuff from each other... it's a curse I know.... you can bet your ass I'd upgrade if it was available

 

[Chip]

A power upgrade will not add a whole lot more, it may attract new users but developers would have to deal with 2 different hardwares.

That's really the key point. This isn't Apple who needs to put out a spec bumped version of each product every 8-12 months just to keep people interested. With the Pandora, it just doesn't make any sense to upgrade the hardware unless the new hardware will bring something new to the table. A dual-core 1GHz A9 Pandora 2 wouldn't do anything the current version can't do. It would just do the same stuff faster. While that would be nice, it's probably not nice enough to justify the cost of developing a new board.

Blue-sky dreaming about new hardware is fun. One of my first posts on this board was along those lines. But if you're looking for realistic answers instead of just "wouldn't it be cool if...", there probably aren't any. Craig and ED have both said they'd like to do a new version when it becomes worthwhile, but nobody knows when that might be, or what hardware options will be available when it is. If there is a Pandora 2 some day, the chip it will run on probably hasn't been announced yet.

 

[Exophase]

That's really the key point. This isn't Apple who needs to put out a spec bumped version of each product every 8-12 months just to keep people interested. With the Pandora, it just doesn't make any sense to upgrade the hardware unless the new hardware will bring something new to the table. A dual-core 1GHz A9 Pandora 2 wouldn't do anything the current version can't do. It would just do the same stuff faster. While that would be nice, it's probably not nice enough to justify the cost of developing a new board.

What is the argument here, is it 1 year vs 2 years? Because craigix has said directly that he wants to do things like Apple does and doesn't intend to fall behind, looking at a revision around 2 years off. Of course, finding the post would be very difficult. Nonetheless, this is nothing more than moving to a dual-core 1GHz Cortex-A9 so obviously he thinks the cost is justified.

 

[anexanhume]

A dual-core 1GHz A9 Pandora 2 wouldn't do anything the current version can't do. It would just do the same stuff faster. While that would be nice, it's probably not nice enough to justify the cost of developing a new board.

What exactly do you mean by that? It couldn't run new applications, or it couldn't run a new class of applications? If the null dc port sees the light of day on the pandora, it's reasonable to assume it may never reach "playable" FPS. However, with more powerful hardware, it arguably could. One could argue that you could also begin considering a port of Yabause given that the psp has seen a port and ARM is becoming more prevalent in netbooks/smartbooks/tablets etc. Those two alone would make me interested in an upgrade.

What is the argument here, is it 1 year vs 2 years? Because craigix has said directly that he wants to do things like Apple does and doesn't intend to fall behind, looking at a revision around 2 years off. Of course, finding the post would be very difficult. Nonetheless, this is nothing more than moving to a dual-core 1GHz Cortex-A9 so obviously he thinks the cost is justified.

Not that difficult:

http://www.gp32x.de/board/index.php?/topic/50804-what-are-the-conditions-for-a-successor-development/page__view__findpost__p__784319

Because if we don't, someone else will.

We have to keep up with technology, look at apple. They do it in a neat way I think.

Or I could bring up Intel...

 

[darionapole]

just a teorical question: how many cortex-A9 cores, at 1 GHz would we need to be able to run a port of pcsx2, let's say playing FinalFantasy X at 45+ FPS ?
fullspeed dreamcast is just a step from posible in the pandora 1, but the next consoles would be saturn, ps2 and gamecube/wii.

what would be needed (in arm) to emulate one of those?

 

[svs-57]

Many people including me very sad about OP formfactor.
Will be nice if OP2 will be slider or better has formfactor like Zaurus sl750 or 1000
Display of Zaurus can be turned 180 degree and put on keyboard.
It can be mini notebook or tablet.

 

[anexanhume]

Many people including me very sad about OP formfactor.
Will be nice if OP2 will be slider or better has formfactor like Zaurus sl750 or 1000
Display of Zaurus can be turned 180 degree and put on keyboard.
It can be mini notebook or tablet.

The slider form factor is patented, and OP is not going to cough up the royalties to use it.

 

[Chip]

What is the argument here, is it 1 year vs 2 years?

No, it's more of "a definite roadmap" vs "whenever it makes sense". Right now, it doesn't make sense, and nobody can predict when it will.

 

[Exophase]

Not that difficult:

Difficult for me anyway. Was it in his last 200 posts? I took a look at thread dates and figured it wouldn't be.

just a teorical question: how many cortex-A9 cores, at 1 GHz would we need to be able to run a port of pcsx2, let's say playing FinalFantasy X at 45+ FPS ?

No number of 1GHz Cortex-A9 cores would allow this. Most emulators don't scale that well beyond a few cores, because some major portion of the runtime has to be ran serially. Actually, most emulators don't even give that much benefit past one core, although this has been changing a bit. An OMAP4 like platform would also be limited in GPU, although 3D emulation does scale better and could possibly be software emulated with dozens of Cortex-A9s, although the sufficient memory bandwidth would probably still be missing.

 

[anexanhume]

Not that difficult:

Difficult for me anyway. Was it in his last 200 posts? I took a look at thread dates and figured it wouldn't be.

I don't know, I just searched "apple" by user craigix and had to look at maybe 4 posts.

 

[Awakening]

Many people including me very sad about OP formfactor.
Will be nice if OP2 will be slider or better has formfactor like Zaurus sl750 or 1000
Display of Zaurus can be turned 180 degree and put on keyboard.
It can be mini notebook or tablet.

The clamshell design is the best for the Pandora. It is the most practical because it allows for the largest possible screen, keyboard and battery without compromising gamecontrols or connectivity. If you want something between a portable console and a netbook then this is the only way to do it well. The Pandora is designed to be unique, not do be as all the other devices.

BTW, the only Zaurus pics I can find are also clamshell.