New Intel IGP Appears In Linux 2.6.30 Kernel

[phoronix]

Phoronix: New Intel IGP Appears In Linux 2.6.30 Kernel

The merge window for the Linux 2.6.30 kernel is now open and Linus has already accepted a horde of new patches for this next quarterly Linux update. Among what has been pulled in so far is the DRM (Direct Rendering Manager) support for a new Intel chipset...

http://www.phoronix.com/vr.php?view=NzE3NQ

[Wyatt]

I, for one, am not certain I agree this is a step in the right direction for netbooks and MIDs. In the event that I bought one, its purpose would be to provide me web browsing and writing from anywhere with >6 hour battery life. Adding higher-power graphics only detracts from that goal, and would seem to me to miss the point of the "Netbook" moniker. Likewise for all the people expecting to play games and watch HD movies on them (if you can, great; but if you can't you have no right to complain: you got exactly what you paid for).

Unless, of course, this is for some decent sub-watt netbook GPU or something. Then I'd say we're getting somewhere.

[Kano]

hd movie on a 1024x600 display you need at least 1024 wide for pal anamorph coded and you want to watch hd... Then you need at least 1280x720 (for 720p).

[clavko]

hd movie on a 1024x600 display you need at least 1024 wide for pal anamorph coded and you want to watch hd... Then you need at least 1280x720 (for 720p).

OK, I would like to know something here. Does HD content take less processing power to decode if it's meant to be shown on a lower resolution screen? I may know little or none about decoding process and optimizations, but wouldn't you still need an 1080p HD capable decoding system to decode a stream before it's been resized to a lower res?

Or did you only say that there's no purpose using hd content on such a machine and that the content itself should previously be encoded to a lower res beacuse of it?

Thanks!

[Kano]

Without external monitor you would not see a diff anyway. There is h264 encoded sd material too, don't know if that runs fully smoothly with atom cpu decoding.

[bridgman]

OK, I would like to know something here. Does HD content take less processing power to decode if it's meant to be shown on a lower resolution screen? I may know little or none about decoding process and optimizations, but wouldn't you still need an 1080p HD capable decoding system to decode a stream before it's been resized to a lower res?

Displaying a video at the native resolution (ie the resolution at which it was encoded) uses the least overall power because you don't have to scale it during the final render stages. AFAIK the docoding part has to run to completion at native resolution no matter what scaling is being done during render.

Scaling down loads the GPU or CPU just like scaling up, although scaling to a bigger screen usually requires more work than scaling to a smaller screen because more pixels have to be written.

[Kano]

Scaling does not matter when you can use xv, but that's not always the case with fglrx

[bridgman]

I was talking about total (CPU and GPU) processing power -- thought that was the question.

[clavko]

Without external monitor you would not see a diff anyway. There is h264 encoded sd material too, don't know if that runs fully smoothly with atom cpu decoding.

No problem, I get that part. The thing is, high definition content is going to get more and more available, pushing the availability of lower resolution content behind. In a situation where SD content is not available or is harder to obtain, it does make sense having an HD capable engine, even if the target screen isn't HD itself. Provided, ofcourse that decoder can't magically reduce complexity of a decoding process to an SD screen (as bridgman indicated).

[portets43]

so could this be the gn40 chipset for the atom?