Pandora OMAP4 and POSITIVE FEEDBACK!!

[second exodous]

I have thought of a device that would have two batteries in it, one main and the other there to give you a few minutes to swap the main out. This would keep clutter down I'd have to take on flights while keeping the device slim.

 

[Exophase]

I have thought of a device that would have two batteries in it, one main and the other there to give you a few minutes to swap the main out. This would keep clutter down I'd have to take on flights while keeping the device slim.

A tiny dongle on the external power should have no problem giving you at least several seconds, if the unit can switch between external and battery power depending on whether or not it senses a cable insertion switches the connection. Maybe could come with a little switch to turn it on/off so you could leave it in. No real reason to need more than half a minute or so, since you just need to do it once you're ready to swap.


The annoying thing would be charging it.


I wonder how much capacitance it'd take to give you a few seconds from capacitors..

 

[Sugar_Kane]

I am utterly convinced that Exophase owns at least one pair of golden underpants

 

[second exodous]

Maybe each individual battery could have it's own mini usb port so it would be possible to charge multiple batteries at once. I wouldn't mind a smaller battery if they were easily switched and didn't need any proprietary cable/dock to charge and USB cables are abundant and cheap to buy if needed. This would also negate the need for a charger to be taken places.


Then again I don't know what adding a mini-usb port capable of charging to each battery would add to the bulk. Maybe that could be the power connector inside the Pandora so at least get a dual use out of it.

 

[gadgetoid]

- any people who are working hard on code for Pandora or meant to eventually be on it will not be happy with you telling them to post less and code more.

Well, I can't exactlty whip them over the internet, can I? Mu ha ha.


Eee. Thread murder.


Anyway. The reason why I'd want an OMAP 4 Pandora can be summed up in one word: Dreamcast.


Although that's not technically a word. I imagine Dreamcast is possible on the current Pandora, but nobody seems to be putting in the effort- I wish I had the cahones to program at that sort of level- hopefully OMAP 4 will, I suppose, relax some of the optimisations required to actually make it playable.


Ultimately, we live in a world where consumer computing, with few exceptions, solves performance problems by chucking ass-tons of hardware into a platform. I'm happy for it to stay that way, as long as the new hardware is sufficiently shiny and/or has blue LEDs.

 

[Gring]

What is the point of having hdmi output on such a small device? An external full hd screen is not bery portable...


Anyway, a Pandora 2 would be mainly a motherboard upgrade, and maybe a few modification to the cases. Now that they went through all the production problems, they could make a new one quickly.


It would take something like 2 months.

 

[TrashyMG]

Insanity Confirmed. Would Vote Again A+++++++++

 

[gadgetoid]

What is the point of having hdmi output on such a small device? An external full hd screen is not bery portable...

Surely you know someone who:


1) Is interested in the Pandora/classic/homebrew multiplayer gaming


2) Owns an HDTV


HDMI output would be awesome, not only for high quality gaming at home, but elsewhere too.


Not to mention the fact that HDMI is basically DVI with some audio thrown in for flavour. DVI out to a desktop monitor would mean that you could achieve enough screen real-estate from a Pandora to, with the addition of a mouse and keyboard, get some *real* work done.

 

[Exophase]

I imagine Dreamcast is possible on the current Pandora, but nobody seems to be putting in the effort

Yeah, you don't really have any idea what you're talking about and should probably stop.

 

[milkshake]

i dont suppose it would be possible for a future version of the pandora to have an esata port? that would be mint. you can also get esata memory sticks now with quite good read write speeds so that would be good.

 

[Exophase]

i dont suppose it would be possible for a future version of the pandora to have an esata port? that would be mint. you can also get esata memory sticks now with quite good read write speeds so that would be good.

Extremely unlikely. The candidate SoCs for Pandora would never have SATA support natively so you'd need an external transceiver hooked up over some kind of high speed interface that may not really exist either, and it'd probably use more power than you would really care to.

 

[milkshake]

Extremely unlikely. The candidate SoCs for Pandora would never have SATA support natively so you'd need an external transceiver hooked up over some kind of high speed interface that may not really exist either, and it'd probably use more power than you would really care to.

SoCs?

 

[Exophase]

System on chip.. the chip on the handheld that has a lot of if not most of the critical functionality. On Pandora that's OMAP3530.

 

[milkshake]

System on chip.. the chip on the handheld that has a lot of if not most of the critical functionality. On Pandora that's OMAP3530.

what about future iterations of the pandora though? omap4?

 

[Exophase]

That's what I'm telling you. OMAP4 doesn't have SATA. SoCs made for phones don't need SATA.

 

[SteveInAustin]

I have yet to get a Pandora console. But I just got a kick out of how it's using the Texas Instruments OMAP chip. I was on the chip design team that designed it. Well, the DSP section anyway. Never thought it would be used for this sort of application. Ha! Can't wait to get my hands on a console.


- Steve


Austin, Texas

 

[WizardStan]

You worked on the DSP? Want to help us get the kernel module built so we can actually use the darn thing?

 

[SteveInAustin]

Hehe. Sorry, I designed a small part of the DSP. I never actually learned to program on it, believe it or not. VLIW architecture is particularly difficult to do by hand in assembly language. Well, easy to do if you don't care to take advantage of VLIW, but then you might as well use the regular ARM core. TI has a proprietary compiler suite that encapsulates a lot of the intelligence in ordering instructions properly. I wasn't privy to the compiler part of the product, or whether or not any of that is open and available to gcc.


- Steve

 

[WizardStan]

You misunderstand, it's not compiling for it that is the problem (although I'm sure it will be eventually), it's actually the bridge to get code onto it: it's a bugger to compile into the kernel, so right now it's just extra transistors that we have no way of accessing

 

[larry]

I expect it'll be several years before we see a new Pandora board. Consider this extremely rough characterization of the rise in power of gaming consoles:

• circa 1995: PS1, N64, tens of MHz
  
• circa 2000: PS2, Gamecube, XBox, Dreamcast: hundreds of MHz
  
• circa 2005: PS3, XBox 360: single-digit gigahertz
  

So, extremely generally speaking, you get a new console every five years, and its CPU is 10x faster. Any faster than that and you risk alienating your user base. (Would you rather buy a new console every five years, or every two years?)


The OMAP 3530 powering the Pandora is a 2009 part, so let's describe the Pandora as a 2009-ish design. If Pandora follows this rough trend, they'll pop out a new board in 2014. (Though perhaps the tireless Open Pandora team will surprise us and deliver a year or two early.)


However, consider this! Once you've got a fixed hardware platform in the hands of developers, over time they learn more tricks and beat better performance out of the rig. Later PS1 games look way better than early PS1 games; later PS2 games look way better than early PS2 games. It's early days yet for Pandora developers. So we should still expect its games and emulators to run better and smoother over time. Who knows, maybe they'll crack DS emulation yet


One thing that I hope happens: we get more asymmetric multicore designs like the Marvell "Armada 628" SOC. That's the three-core ARM they announced in September. Two cores are (up to) 1.5GHz, the third is 624MHz and much lower power. They claim it can "handle 140 hours of music on one charge"; presumably that's with the graphics processor and the two high-speed cores shut down. Just think: in the future, we could get a Pandora that could emulate PS2 level consoles, yet also when used for simple computing tasks (email/web/music) could run several times longer on a charge.