Boxee drops tegra 2?
#16
Why did they have to make it so weird looking? I absolutely hate the odd shape of it. I may get one just because it seems really wife friendly but I'd of preferred just a normal dvd-player type design. Sad to hear about the tegra 2 as I was hoping the next round of tablet PCs with tegra 2 and hdmi out would be win/win.
Reply
#17
Robgue Wrote:I'm getting one of these. I was worried about the Tegra 2 but I'm for sure getting it now. I would think especially now that it's not arm based and boxee understanding people want some openness running xbmc won't be too far down the pipeline.

Agreed, being x86 - this new Intel CPU could be the new hope! Smile

Also see that ASUS have a box coming out later on using the same CE4100 CPU

http://www.engadget.com/2010/09/13/asus-...ter-embed/

Running WMCE it's x86 as well.

Sort of makes sense why Intel stunted ION2 like they did if they had this as their new 'media player' CPU in the works.
Reply
#18
davilla Wrote:Forget about GPU doing video decoding on these type of platforms, there is a dedicated video decoder present on the chipset.

So it sounds like we are looking at a box like Popcorn hour or Western digital box.

Is it safe to say that the intel board they are using has a sigma or realtek video decoder? Or intels first run at a video decoder? I know we just heard today about the jump, but I figured I would ask.
Reply
#19
BORIStheBLADE Wrote:So it sounds like we are looking at a box like Popcorn hour or Western digital box.

Is it safe to say that the intel board they are using has a sigma or realtek video decoder? Or intels first run at a video decoder? I know we just heard today about the jump, but I figured I would ask.

http://www.anandtech.com/show/3912/boxee...side-story

They claim it's likely a Imagination Technologies PowerVR series decoder which would match the GPU.
Reply
#20
Looking at the block diagram for the CE4100, the Multi-Format HW Decoder, Display Processor and Graphics Processor are all inside a block. This suggests they might be related. Since only one has any number ( SGX535 ), this suggests Imagination Technologies is involved.
Reply
#21
That is the same GPU used in Intel "Paulsbo" GMA500 chipset. Maybe this means that they are going to have a better linux driver for it?
Reply
#22
I read that the Boxee Box will run Intel's Linux build so I suppose it will have all the driver optimizations needed.
Reply
#23
GMA500 has closed source Vaapi support, mostly just for Ubuntu (certain versions):

https://wiki.ubuntu.com/HardwareSupportC...dsPoulsbo/

Reply
#24
Hopefully it will be possible to install XBMC alongside boxee on this thing using the same internal linux distribution. There is the question of whether the internal storage has enough free space to throw XBMC on as well or if a USB stick will have to be put in place to make supplemental space available.
Subtitles - Serious Business
Reply
#25
So does this box support fanarts?
Reply
#26
russg Wrote:I read that the Boxee Box will run Intel's Linux build so I suppose it will have all the driver optimizations needed.

I guess, this is great news for XBMC users regarding Intel GPU's and AMD/ATI (porting back used video-drivers and VAAPI support) and flash 10.1 playback in Linux....
Reply
#27
Robotica Wrote:I guess, this is great news for XBMC users regarding Intel GPU's and AMD/ATI (porting back used video-drivers and VAAPI support) and flash 10.1 playback in Linux....

....if that stuff gets opened up. You never know what things the Boxee people agreed to in order to finally ship this thing.

Intel's GMA500 driver was closed source (or had closed parts) prior to the Boxee box, and something tells me the Boxee folks don't have enough swing to force Intel to force PowerVR to allow Intel to open up that driver.

Considering how the Boxee box changed from Tegra to something else "at the last minute," I feel like we will be lucky if this thing really has a open source base (aka at a driver level).

You never know- maybe backed into a corner after choosing initially an underwhelming platform the Boxee folks maybe had to accept some value compromises to actually get a product out the door. Happens all the time.

I am not saying its a guarantee that the Boxee box is filled with locked down tech, but at the same time I think its dangerous for everyone to assume that since Boxee is involved (and Boxee is a supportive daughter project of XBMC) the whole thing will be open and we will have XBMC on there in a week.

Usually anything involving Flash is not open, and as I said that Intel driver is not open. Someone like IBM, Dell, or HP have the leverage to demand opneness- Boxee lacks that leverage. They needed the cheapest box possible that would play Boxee, and that might be a closed box. The other cheapest boxes in the market are.

This box is it- Boxee's shot. If this thing fails, Boxee fails.

Who knows what they have agreed to (especially after a last minute hardware switch) to make sure they actually got their shot?

We will soon know.

Reply
#28
It would be easy to wrap Intel's closed source bits into a OpenMax IL decoder component. That's an Open API which XBMC can use to access hardware video decode. A similar thing is done under tegra2 with Nvidia's closed source bits.

Almost all embedded SoCs that use EGL for display handling have gone the OpenMax route for hardware decode handling.
Reply
#29
davilla Wrote:Almost all embedded SoCs that use EGL for display handling have gone the OpenMax route for hardware decode handling.

If that is the case, then the only thing that might come locked down is if it has some special version of Flash.

Its gonna be crazy in here in a few months!!!!!

Reply
#30
Off Topic but:

http://labs.adobe.com/downloads/flashplayer10.html

Skunkworks:chd-test files davilla$ strings /Users/davilla/Downloads/libflashplayer.so | grep libcrystalhd
libcrystalhd.so

Bruhaha !!!!
Reply

Logout Mark Read Team Forum Stats Members Help
Boxee drops tegra 2?0