How About The 720Mhz Omap For The Second Batch?

[Ollonk]

Confirmed: Pandora 2 will be in our hands in two months.

 

[Exophase]

You really should not worry. The Pandora2 will almost certainly be an OMAP4 and judging by the OMAP3 time line that's not going to be ready until at least the end of 2011.

So maybe you could see some dev units then and a push in mid 2012 for customers.

In other words, it's 2 months away.

I think you're going to be eating your words if an OMAP4 device comes out over a year earlier than you project. If TI doesn't plan on having OMAP4 ready to market in 2010 then they'd probably may as well just exit the market - Tegra 2 can be obtained today. NEC and NXP have also both announced Cortex-A9 SoCs. With A9 being announced all the way back in 2007 my guess is that some companies skipped out on A8 altogether.

I wouldn't judge the OMAP4 release cycle based on OMAP3, which was still the first Cortex-A8 SoC to market after all the time it took to release. TI did claim it'd be sampling in 2010.

 

[Uh...Clem]

You really should not worry. The Pandora2 will almost certainly be an OMAP4 and judging by the OMAP3 time line that's not going to be ready until at least the end of 2011.

So maybe you could see some dev units then and a push in mid 2012 for customers.

In other words, it's 2 months away.

I think you're going to be eating your words if an OMAP4 device comes out over a year earlier than you project. If TI doesn't plan on having OMAP4 ready to market in 2010 then they'd probably may as well just exit the market - Tegra 2 can be obtained today. NEC and NXP have also both announced Cortex-A9 SoCs. With A9 being announced all the way back in 2007 my guess is that some companies skipped out on A8 altogether.

I wouldn't judge the OMAP4 release cycle based on OMAP3, which was still the first Cortex-A8 SoC to market after all the time it took to release. TI did claim it'd be sampling in 2010.

Looks like TI will be releasing an amazingly capable development platform for the OMAP4 in the near future which I think will go for US$500:

http://www.youtube.com/watch?v=z0STvlyCxog

Another awesome near-term SoC, the Marvell Armada 618:

http://www.youtube.com/watch?v=pKBibQgLaVY

 

[quadomatic]

You really should not worry. The Pandora2 will almost certainly be an OMAP4 and judging by the OMAP3 time line that's not going to be ready until at least the end of 2011.

So maybe you could see some dev units then and a push in mid 2012 for customers.

In other words, it's 2 months away.

I think you're going to be eating your words if an OMAP4 device comes out over a year earlier than you project. If TI doesn't plan on having OMAP4 ready to market in 2010 then they'd probably may as well just exit the market - Tegra 2 can be obtained today. NEC and NXP have also both announced Cortex-A9 SoCs. With A9 being announced all the way back in 2007 my guess is that some companies skipped out on A8 altogether.

I wouldn't judge the OMAP4 release cycle based on OMAP3, which was still the first Cortex-A8 SoC to market after all the time it took to release. TI did claim it'd be sampling in 2010.

Certainly seems like OMAP4 isn't too far away:

http://www.engadget.com/2010/02/15/texas-instruments-introduces-arm-based-omap-4-soc-blaze-develop/

the guy recording the video that clem posted for the OMAP4 is really stupid....

 

[Exophase]

Those Blaze videos are awesome, thanks. I never even knew this dev kit was coming. OMAP4 looks like it's going to be quite a bit nicer than Tegra 2.

Man I want a Blaze now.

 

[Tensuke]

Those Blaze videos are awesome, thanks. I never even knew this dev kit was coming. OMAP4 looks like it's going to be quite a bit nicer than Tegra 2.

Man I want a Blaze now.

I would give up my firstborn for one.

 

[relliker]

so, craigix actually made a TWO MONTHS joke. Unbelievable! Makes you almost believe that he's human.

 

[craigix]

You really should not worry. The Pandora2 will almost certainly be an OMAP4 and judging by the OMAP3 time line that's not going to be ready until at least the end of 2011.

So maybe you could see some dev units then and a push in mid 2012 for customers.

In other words, it's 2 months away.

I think you're going to be eating your words if an OMAP4 device comes out over a year earlier than you project. If TI doesn't plan on having OMAP4 ready to market in 2010 then they'd probably may as well just exit the market - Tegra 2 can be obtained today. NEC and NXP have also both announced Cortex-A9 SoCs. With A9 being announced all the way back in 2007 my guess is that some companies skipped out on A8 altogether.

I wouldn't judge the OMAP4 release cycle based on OMAP3, which was still the first Cortex-A8 SoC to market after all the time it took to release. TI did claim it'd be sampling in 2010.

Being able to buy it does not mean it's ready. The software needs to be mature too. I think it will take just as long as the OMAP3 has.

 

[WizardStan]

Being able to buy it does not mean it's ready. The software needs to be mature too. I think it will take just as long as the OMAP3 has.

And if you're wrong, and it is a shorter time to market, it's reasonable that some OMAP5 can't be that far off either. Maybe we just skip a generation and get the bleeding edge then.

 

[rv6502]

they announced the iPhone 3GS a month after I bought a 3G... boohoo.

news flash: CPUs and devices get faster and faster with newer generations.

You bought an iPhone 3G. They then created an entirely new device (which looked and acted very similar) to that called the 3GS. They didn't bump up the specs of an existing device.
Any decision to change the Pandora, adding extra features or a new CPU, will result in a Pandora 2, an entirely new device which will look and act very similar to the original (indeed, Craigix said they're hoping to keep the same case mould, or at least design, when they're ready to create a new version)
Putting a new CPU into the second batch and still calling it Pandora would be as much a slap in the face as putting all the extra stuff that went into the 3GS and still calling it just a 3G. Not only because it alienates early adopters, but more importantly it makes it difficult to maintain a standard platform. Developers would have to differentiate between "3G" and "3G with extra stuff" somehow, right?

edit: corrected an incorrect double negative

the point I'm trying to make: some people on the forum whine about how late adopters might be getting better deals due to the normal flow of technological improvements in chip production.

that's just life, we're not gonna stop progress because of some kinder-garden mentality.

if for some reason TI stops making 600Mhz parts and/or all parts OC better and the OP team has to produce using >600Mhz parts for the 2nd, 3rd or whatever batch and if people in the 1st batch want to feel cheated about it, s to be them.

I'm getting (someday ... maybe in two months :lol: ) a 1st batch Pandora before all the 2nd, 3rd, 4th, etc, batchers and if later batches run faster that will be the price I paid to get mine before everyone who came after. be it 4 years later, half a year later, one month later, or even two days later.


@craigix: I got an idea, why not offer a Pandora upgrade service for whiny early adopters:
"send back your 1st batch Pandora and for a low-low price of just 330$ [sub](+tx, shipping and handling)[/sub] we'll upgrade it to the latest"

 

[Exophase]

Being able to buy it does not mean it's ready. The software needs to be mature too. I think it will take just as long as the OMAP3 has.

I dunno, it looks like it's in pretty good shape already. I don't think there's that much on OMAP4 that'll be a big migration challenge over OMAP3. Multicore isn't really a big deal.

 

[Usoppe]

Being able to buy it does not mean it's ready. The software needs to be mature too. I think it will take just as long as the OMAP3 has.

And if you're wrong, and it is a shorter time to market, it's reasonable that some OMAP5 can't be that far off either. Maybe we just skip a generation and get the bleeding edge then.

Iam no economist but I see several reasons to skip the OMAP4:

a ) OMAP4 devices should be avaible in about -6month I think, so if the Pandora 2 bases on it too, it will be ready in 2 years probably (1,5 year development, production + the team would start after the delivering of lets say second batch of Pandora 1, so 0.5 years). But then it will be old gen -> bad for your sales.
b ) Ask TI in about 1 years about the OMAP5 and that you want to use this in your next gen device. Then you work for > 1.5 year to make a Pandora 2 ready - and the most complicated part would be the board design, I think. The Pandora 2 should be ready in 2.5, 3 years.
c ) As said, if you dont skip the OMAP4 you could deliver a new Pandora in perhaps ~2 years. But its too short if you ask me and it would be bad for the sales of the Pandora 1 (newer device avaible) and the Pandora 2 (OMAP4 is old stuff again) . Better wait 1 more year with the Pandora 2 and go 2 steps to the OMAP5 because you wont be so far to cut-edge technology (as a small company, it would be a big achieve).
d ) Waiting has another advantage because alot of software will be avaible, too. You can better market the Pandora 2 then. And I think that its not very complicated to port a program for an OMAP3 chip to future ones (except they change the ARM core significantly, but most probably future cores will be downwardy compatible).

Just my thoughts.

Here some interesting rumors about the OMAP5. They say that the PSP2 could base on it:
http://www.brightsideofnews.com/news/2009/10/16/exclusive-sony-psp2-arriving-in-20102c-features-powervr-sgx543-graphics.aspx
http://www.eurogamer.de/articles/digitalfoundry-ds2-vs-psp2-article_1?page=2 (german)

Last news rumors that the PSP2 wont be avaible before 2011. Lets say it will appear at the end of 2011 at xmas time. So about 1,5 year from now. Wouldnt it be great if the Pandora 2 appears with a OMAP5 just 1, 1.5 years after the PSP2? Then you can claim that youre as fast as the PSP2 but open source

Another thing: Its better to compete against hand-held consoles instead of mobile phones - mobile phone are NO competition in my eye - no gaming controls and not a real gaming device - casual games like "Snake 2.0" (synonym for Iphone games) on it, not really competetive. The Pandora 1 is good placed, right now, theres no hand-held console which is faster.

But it wont look good for the Pandora 2 (with OMAP4) when putting it on the market when the PSP2 (with OMAP5 or similiar) appears. Wait a bit longer, make a Pandora 2 with an OMAP5, be in 1, 1.5 year after the PSP2 on the market and have about 3-4 years time after delivering the Pandora 2 (time before a PSP3 will appear) to develop a future Pandora 3.

That sounds good to me .

 

[torpor]

Just give us the Pandora, now! Whatever the CPU/RAM, we developers will make it shine.

 

[mali]

@Usoppe
Skipping OMAP4 would be a bad idea, imo. Things never work out as expected, so if the gap between iterations is too large, the community might migrate elswhere. You need an upgrade path perspective on the horizon to tide the users over.

 

[Usoppe]

[Deleted post, updated my post before instead. Updated it second time..]

 

[Aninhumer]

I think arguing about what hardware to use in a product which is at LEAST 2 years away, is pretty pointless.
A LOT can happen in two years.

 

[mali]

^ I agree. The only important premise is, that it should be fairly downward compatible.

 

[Usoppe]

But you guys know that it needs to be developed and it takes time ? But youre right if you say that we should start talking in about one year.

 

[Uh...Clem]

TI booth demos 65nm OMAP3430 vs 45nm OMAP3630:

http://www.youtube.com/watch?v=lQQM3w8atsM

Pandora is mentioned by the interviewer very near the end of the video.

 

[diegor]

You really should not worry. The Pandora2 will almost certainly be an OMAP4 and judging by the OMAP3 time line that's not going to be ready until at least the end of 2011.

So maybe you could see some dev units then and a push in mid 2012 for customers.

In other words, it's 2 months away.

I think you're going to be eating your words if an OMAP4 device comes out over a year earlier than you project. If TI doesn't plan on having OMAP4 ready to market in 2010 then they'd probably may as well just exit the market - Tegra 2 can be obtained today. NEC and NXP have also both announced Cortex-A9 SoCs. With A9 being announced all the way back in 2007 my guess is that some companies skipped out on A8 altogether.

I'm not sure that you can buy now. It's quite standard to announce some new product to be available "real soon" and then releasing after 12-18 months. Expecialy if their current product is inferior

And on the nvidia site, tegra2 is under "next generation tegra", but with any "buy" link. So probably is not ready at all.

edit: deleted some unmatching tag