Looking For Partners To Finish My Tablet

[00christian00]

DELETED

 

[DaveC]

Sounds interesting as the Pandora case designer. The thing is though there is a flood of droid tablets coming judging by the CES reports. What will make yours stand out from the pack?

Unless yours has something unique in it like a lenticular 3D screen or OLED screen you would be lost in the shuffle.

 

[00christian00]

Yes,that announcement also scared me.Anyway to answer you:
1-As I said above,WHDI,FULL HD wireless transmission.Although I read on Engadget that they are planning to announce some tablet with this,I didn't found any actual report from any website,so it seem a little strange.
Plus my Amimon source delayed the transmitter until 1 week ago,so I doubt they really had time to integrate it in some product already for CES.
2-Video playback.Most of the tablet are based on Tegra 2,and I remember Boxee ditched it because it had problem with high bitrates.
Plus as far as I know Tegra 2 doesn't have that many codec supported,while Ziilabs already got WebM/VP8 supported already.
3-OpenCL.The device could be targeted to some custom application,and being able to use OpenCL will mean that the device can outperform easily any dual core soc out now,since it have 8 cluster of 8 units running at 200mhz.
http://www.youtube.com/ziilabs

 

[rabidpoobear]

Yes,that announcement also scared me.Anyway to answer you:
1-As I said above,WHDI,FULL HD wireless transmission.Although I read on Engadget that they are planning to announce some tablet with this,I didn't found any actual report from any website,so it seem a little strange.
Plus my Amimon source delayed the transmitter until 1 week ago,so I doubt they really had time to integrate it in some product already for CES.
2-Video playback.Most of the tablet are based on Tegra 2,and I remember Boxee ditched it because it had problem with high bitrates.
Plus as far as I know Tegra 2 doesn't have that many codec supported,while Ziilabs already got WebM/VP8 supported already.
3-OpenCL.The device could be targeted to some custom application,and being able to use OpenCL will mean that the device can outperform easily any dual core soc out now,since it have 8 cluster of 8 units running at 200mhz.
http://www.youtube.com/ziilabs

I guess the problem with this SoC then is that things must be written specifically to perform well on it; this is the same thing PS3 went through with Cell, except Cell has many millions of installed base and software for PS3 sells for $60 a pop - a mobile device like a Pandora 2 will not have nearly as much of a market share and thus fewer developers - it becomes very important then to have our next processor be faster at *existing* software, not at *theoretical* software. Because, if we're honest, most Pandora software is ports and surely will be that way for quite a while into the Pandora team's future.

Just my 2 cents as a developer. I don't want to learn a new architecture and a new paradigm of programming efficiency that will only apply to a single platform.

 

[snostrebla]

I stopped reading after I saw two months. Joking of course.

Best of luck on your venture, just by posting here I can imagine everyone here knows how hard it is to build something the world will appreciate.

I can't personally help, and I don't know anywhere else one could post this, I'm sure there are websites out there for this purpose though.

 

[00christian00]

Yes,that announcement also scared me.Anyway to answer you:
1-As I said above,WHDI,FULL HD wireless transmission.Although I read on Engadget that they are planning to announce some tablet with this,I didn't found any actual report from any website,so it seem a little strange.
Plus my Amimon source delayed the transmitter until 1 week ago,so I doubt they really had time to integrate it in some product already for CES.
2-Video playback.Most of the tablet are based on Tegra 2,and I remember Boxee ditched it because it had problem with high bitrates.
Plus as far as I know Tegra 2 doesn't have that many codec supported,while Ziilabs already got WebM/VP8 supported already.
3-OpenCL.The device could be targeted to some custom application,and being able to use OpenCL will mean that the device can outperform easily any dual core soc out now,since it have 8 cluster of 8 units running at 200mhz.
http://www.youtube.com/ziilabs

I guess the problem with this SoC then is that things must be written specifically to perform well on it; this is the same thing PS3 went through with Cell, except Cell has many millions of installed base and software for PS3 sells for $60 a pop - a mobile device like a Pandora 2 will not have nearly as much of a market share and thus fewer developers - it becomes very important then to have our next processor be faster at *existing* software, not at *theoretical* software. Because, if we're honest, most Pandora software is ports and surely will be that way for quite a while into the Pandora team's future.

Just my 2 cents as a developer. I don't want to learn a new architecture and a new paradigm of programming efficiency that will only apply to a single platform.

It's OpenCL,It's a standard for Desktop parallel computing, and will become soon the standard for mobile too,once other vendor will incorporate it.All the new mobile GPU should have it.

 

[rabidpoobear]

It's OpenCL,It's a standard for Desktop parallel computing, and will become soon the standard for mobile too,once other vendor will incorporate it.All the new mobile GPU should have it.

I misunderstood you, the way you compared it to the dual core SoC's it soudned like you were discussing the processor architectures, not the GPUs.

I don't really see the advantage then; won't Tegra have OpenCL soon? NVIDIA has done it on all their desktop GPUs already. And it doesn't seem like the Zii is available for large scale acquisition yet, or am i wrong about that?

 

[00christian00]

It's OpenCL,It's a standard for Desktop parallel computing, and will become soon the standard for mobile too,once other vendor will incorporate it.All the new mobile GPU should have it.

I misunderstood you, the way you compared it to the dual core SoC's it soudned like you were discussing the processor architectures, not the GPUs.

I don't really see the advantage then; won't Tegra have OpenCL soon? NVIDIA has done it on all their desktop GPUs already. And it doesn't seem like the Zii is available for large scale acquisition yet, or am i wrong about that?

Maybe Tegra3,they don't have it in Tegra 2.What do you mean by "large scale acquisition"?

 

[rabidpoobear]

Maybe Tegra3,they don't have it in Tegra 2.What do you mean by "large scale acquisition"?

I mean if you wanted to build a device using them... could you purchase enough? it sounds like it's still in prototype and design stage. I mean it's all nice for them to promise performance but if you can't even buy the chips then what good is it?

Re: openCL, isn't it essentially just a different way of using shader pipelines? couldn't they write a driver for the current GPUs in the Tegra 2? I mean they already have the shader architecture in place on their systems to support OpenGLES.

 

[00christian00]

Yes,it can be bought.The cost is much less than omap3530 and the lead time is 8 weeks.I know for sure it can be bought as I am in close cooperation with them.

OpenCL is for general computing,not for shaders.
Basically you run normal software on the GPU inside the CPU.Yes they could do it,but it's not easy thing and they didn't do it yet.And the tegra 2 gpu doesn't seem that powerful anyway.

 

[rabidpoobear]

Yes,it can be bought.The cost is much less than omap3530 and the lead time is 8 weeks.I know for sure it can be bought as I am in close cooperation with them.

OpenCL is for general computing,not for shaders.
Basically you run normal software on the GPU inside the CPU.Yes they could do it,but it's not easy thing and they didn't do it yet.And the tegra 2 gpu doesn't seem that powerful anyway.

Yeah I researched it and it sounds from the nvidia engineer's comments that I've read that they don't expect it to ever come to Tegra 2.

Well 8 weeks is nice, and volumes would probably be pretty low - what parts do you not have finished yet? You have already elected which touchscreen / LCD you'll use and all, do you just need a case? or what all is missing?
Is your PCB very high in layer count or require any complicated production (i.e. would the assembly cost more / have a lower yield due to complexity of the PCB? or is it low enough that you know of a Shenzhen manufacturer that could make these cheaply w/ high yield)

[edit] the reason I mentioned shaders specifically is that that was essentially what they were demoing the unit doing in that youtube link you posted - applying effects to images, raytracing, etc. all can be done on shaders. [/edit]

 

[00christian00]

Yes,it can be bought.The cost is much less than omap3530 and the lead time is 8 weeks.I know for sure it can be bought as I am in close cooperation with them.

OpenCL is for general computing,not for shaders.
Basically you run normal software on the GPU inside the CPU.Yes they could do it,but it's not easy thing and they didn't do it yet.And the tegra 2 gpu doesn't seem that powerful anyway.

Yeah I researched it and it sounds from the nvidia engineer's comments that I've read that they don't expect it to ever come to Tegra 2.

Well 8 weeks is nice, and volumes would probably be pretty low - what parts do you not have finished yet? You have already elected which touchscreen / LCD you'll use and all, do you just need a case? or what all is missing?
Is your PCB very high in layer count or require any complicated production (i.e. would the assembly cost more / have a lower yield due to complexity of the PCB? or is it low enough that you know of a Shenzhen manufacturer that could make these cheaply w/ high yield)

[edit] the reason I mentioned shaders specifically is that that was essentially what they were demoing the unit doing in that youtube link you posted - applying effects to images, raytracing, etc. all can be done on shaders. [/edit]

The pcb is 8 layers because it is including wifi/bt/gps and 3g(via mini pcie).I already know cheap and fast(relatively cheap,a pcb of this kind is never cheap) for this and also for the smt.
The components right now are the same as the reference design,because this way I don't need to touch the software at all for the testing(otherwise it would be harder to find the hw issues).But after the initial prototype I would like to change the lcd and touch.
Since the interface is LVDS,there would be no hardware change.
The main issue now is that I don't trust the team i created anymore,they are overpaid,don't have the skills they said to have and always argue for everything never taking any responsability...
And I don't want to hire new engineers in China,because the level seem pretty low and the attitude quite bad,probably because I'm a foreigner or they are used to work in another way.The second reason is due to cost,if the team mates are partners a delay does cost only time,if it's employees it cost time and lots of money.

 

[skeezix]

Maybe I'm in a poopy mood, and I feel bad for writing it, but it sure seems someone has to drop in the bomb .. I read it as..

.. I screwed up a lot (and I mean, dude, _everyone_ knows the story of the guy who goes to China and gets screwed like that .. thats what happens there
.. I spent all my money and need someone elses
.. It was a lot
.. I want to enter a market against the biggest guys in the industry without something to differentiate

Sounds very bad to me (especially in light of OPT .. they're not rolling in money, why woudl you ask them to help bail you out?)

Pandora works because of the rare gaming controls on linux clamshell ... And a prebuilt tight knit community who hld on for years bcause they know all the devs

I dunno .. Theres a lot of stories like yours and thats not good

To your credit, you have popped up here a few times and if you moved to China then you are hardcore, but this board is where guys have struggled like hell and they dont have cash to risk.. Hey're running on preorders, not profit. Its a long way before OPT has money, and you're a bad risk

Jeffphone

edit; I thnink you would be better off finding folks who had lotds of expertise in production, and lots of cash. Unfortuantely its tough since you've already botched a few things, they'd be better off on risking with something else (there are hundreds of teams working on tablets right now, with similar stories to yours.)

Stuff like video playback etc is not a big deal; WebM etc is also not really used much these days, its a future thing; you need to worry about current popular codecs such as for yutube, dvd decoding, etc. Remember Tegra 2 can do H264 high def right out of the box, and at 1ghz models can decode dvd's in software not half bad, but also has acceleration for it.

 

[DaveC]

Also trying to come out with something and expecting it to sell because "It uses the most powerful SoC EVAR!" Is a recipe for disaster. By the time your product comes to market it will already be obsolete power wise. Everyone else will have equal power and new chip announcements of even better performance will be known and steal any thunder you had. Then you have the fact that Android kind of sucks and is a resource hog. So in that case you need some kind of feature no one else has to make it desirable. Yet another Google powered tablet will honestly fail. It is not an iPad and the support for your hardware will be low. People don't care what GPU CPU or whatever geektastic tech you may have under the hood, they want features and apps. Right now the Apple devices have that pretty much locked.

As I said you need some kind of exclusive feature no one else has. The two I mentioned would do it but you would need a hardware source. An OLED screen would be great as no tablet has that. Also an autostereoscopic 3D screen would fit a good niche. There are 3D digital cameras on the market with no good solution to display the pictures other than the camera itself (screen is rather small) or a 3D TV. With the introduction of the 3D DS, 3D will get allot of buzz. Again the DS has a small screen so a larger tablet would have a good market. A good 3D tablet would be the ultimate 3D photo and 3D movie PMP viewer and 3D is becoming quite a niche. Maybe look into that.

I could design up a tablet case pretty easily. A few hours in my CAD program and I could pop it out. That would be super simple after the Pandora case. There would have to be some promising feature though to make it worthwhile though.

 

[Esn]

Just to head off any concerns before they are raised, DaveC is who he says he is.

 

[00christian00]

Also trying to come out with something and expecting it to sell because "It uses the most powerful SoC EVAR!" Is a recipe for disaster. By the time your product comes to market it will already be obsolete power wise. Everyone else will have equal power and new chip announcements of even better performance will be known and steal any thunder you had. Then you have the fact that Android kind of sucks and is a resource hog. So in that case you need some kind of feature no one else has to make it desirable. Yet another Google powered tablet will honestly fail. It is not an iPad and the support for your hardware will be low. People don't care what GPU CPU or whatever geektastic tech you may have under the hood, they want features and apps. Right now the Apple devices have that pretty much locked.

As I said you need some kind of exclusive feature no one else has. The two I mentioned would do it but you would need a hardware source. An OLED screen would be great as no tablet has that. Also an autostereoscopic 3D screen would fit a good niche. There are 3D digital cameras on the market with no good solution to display the pictures other than the camera itself (screen is rather small) or a 3D TV. With the introduction of the 3D DS, 3D will get allot of buzz. Again the DS has a small screen so a larger tablet would have a good market. A good 3D tablet would be the ultimate 3D photo and 3D movie PMP viewer and 3D is becoming quite a niche. Maybe look into that.

I could design up a tablet case pretty easily. A few hours in my CAD program and I could pop it out. That would be super simple after the Pandora case. There would have to be some promising feature though to make it worthwhile though.

Sorry but your point is flawed,extremely flawed.
First I never said I have the most power soc ever.They ask me the difference,and I pointed them out.That's all.
Second what you said about being obsolete doesn't apply to soc only,but everything.You says there are no 3d screen and AMOLED tablet now.What make you think that by the time you release the product they won't become mainstream?
Do you know the reason why there are no product using them?Because you cannot buy the screen itself!
Most of the time the company keep it for themselves(see Sharp and Samsung) and what you can buy is like one generation behind them,at least,and at crazy high price.
An AMOLED screen cost up to 3 times a good lcd screen,and now that they even incorporate on cell capacitive sensor,it must be much more....
Good luck trying to sell a small brand tablet at even higher prices than big brands,just because it has AMOLED(which anyway you can't buy at this tablet size yet)...
Then maybe one week later samsung itself announce the galaxy tab 2,at half price than yours because they can get the screen at manufacturing cost.

 

[Caine]

Another tablet? In my opinion, there are already way too many out there. I don't really see the point of the devices anyway. They are too bulky to carry around in your jacket and a laptop has them fully covered in functionality.

So, please enlighten me. Why would I want your tablet?

• work, needs a keyboard/mouse. Laptop is better.
• reading books. E-ink is a must in my opinion.
• gaming, requires buttons
• music, too big
• movies. Maybe one of the valid use-cases, but I seldom watch movies while traveling.

 

[Alibobar]

I am still waiting on that ultra cheap tablet from india. If it ever comes out an indian friend of mine will buy one for me

Personally I don't have any particular use for a tablet either. I think they are currently more novelty items than in any way really functional and handy devices.

So I really hope you don't end up in that endless hole of android-tablet copies and find some really original implementation of your device.

 

[DaveC]

Sorry but your point is flawed,extremely flawed.
First I never said I have the most power soc ever.They ask me the difference,and I pointed them out.That's all.
Second what you said about being obsolete doesn't apply to soc only,but everything.You says there are no 3d screen and AMOLED tablet now.What make you think that by the time you release the product they won't become mainstream?
Do you know the reason why there are no product using them?Because you cannot buy the screen itself!
Most of the time the company keep it for themselves(see Sharp and Samsung) and what you can buy is like one generation behind them,at least,and at crazy high price.
An AMOLED screen cost up to 3 times a good lcd screen,and now that they even incorporate on cell capacitive sensor,it must be much more....
Good luck trying to sell a small brand tablet at even higher prices than big brands,just because it has AMOLED(which anyway you can't buy at this tablet size yet)...
Then maybe one week later samsung itself announce the galaxy tab 2,at half price than yours because they can get the screen at manufacturing cost.

Ok so you are saying that because someone *might* produce a 3D tablet sometime someday (none have been announced) that it is just better to come out with a tablet similar to what HAS already been announced by SEVERAL companies? That makes even less sense. I would rather try to compete with a "maybe someday" instead of a many for sure soon.

Anyway if you are clever you can make your own 3D screen by inserting a parallax barier or lenticular sheet over the LCD internally. Yes it would take a bit of R&D to match the dot pitches but that is the point right? offer something that you can't already get? BTW they have an add-on but it wouldn't be as good as an internal one (can be closer to the LCD panel internally giving better results, plus be protected from wear): http://grilli3d.com/GRilli.html.

 

[Aurekana]

Or else, seeing as it is a 10" tablet, attack the iPad head on with a fix for some of the most common complaints about the iPad that Apple is likely to never fix, as in, make a 10" tablet with physical gaming controls and a USB port.