Linked by Thom Holwerda on Tue 6th Jan 2009 09:36 UTC, submitted by caffeine deprived
Hardware, Embedded Systems It seems that after Intel, just about every chip maker wants a piece of the netbook pie. AMD is an obvious competitor, but VIA is also eyeing the little notebooks. However, more exotic options like the Chinese Loongson chips and ARM's Cortex A-8 and A-9 chips are also among the contenders. We can now add a new contender: Freescale.
Thread beginning with comment 342498
To view parent comment, click here.
To read all comments associated with this story, please click here.
RE[7]: x86 is brain dead
by BluenoseJake on Tue 6th Jan 2009 18:54 UTC in reply to "RE[6]: x86 is brain dead"
BluenoseJake
Member since:
2005-08-11

it's a bit different. A console is not a netbook, even a netbook is supposed to have some of the flexibility of a general purpose computing device.

Also, I modded my old xbox to the hilt, ran Debian on it, and used it as a media center, a webbrowser, and a game console. I enjoyed the fact that it DID have some capabilities to be changed, even if it was against MS's wishes. And that was made easier by the fact that it was a x86.

You can get a couple of distros for the playstation, but I think (and I may be wrong) that it is a lot less compatible with the great majority of Linux distros, because it is either MIPS (ps2) or Cell (ps3)

The original xbox is almost off the shelf PC parts, so it may be one of the most versatile consoles ever, imo

Reply Parent Bookmark Score: 3

RE[8]: x86 is brain dead
by spiderman on Wed 7th Jan 2009 06:51 in reply to "RE[7]: x86 is brain dead"
spiderman Member since:
2008-10-23

At the moment, x86 is more versatile because you can install Windows only on it and most of its software. The problem is that the architecture is very slow for high performance 3D computation. Indeed, even on a desktop PC, you have to add some GPU from NVidia or from ATI to play games with acceptable performances, even with MMX and SSE, the performance still sucks. Indeed, the cell processor may not run Windows, but it can do 3D computations and much more. For instance when you compare opencv performance on a cell processor and on an x86 processor, the cell processor is 27 times faster at the same clock speed, even when x86 has the IPP instruction set enabled. Note that opencv is a library designed by intel to run fast with the IPP instruction set.
Actually, x86 is very bad design and is too slow and expensive for consoles which require high performance 3D computation.
On the netbooks, it is the same. We currently use x86 because only intel is capable of bulding processors with that small transistors that consume that few energy. Running Windows is a bonus, but imagine that intel was dumping Windows compatibility and was designing ARM processors. Imagine a netbook that you could use during 10 hours without plugging. x86 is just for legacy software, but I'm convinced it is not the best architecture for the netbooks and it is now the time to dump this architecture before all netbook software require it. It is now or never. I hope it is now.

Edited 2009-01-07 06:52 UTC

Reply Parent Bookmark Score: 2

RE[9]: x86 is brain dead
by Treza on Wed 7th Jan 2009 13:45 in reply to "RE[8]: x86 is brain dead"
Treza Member since:
2006-01-11

I guess you include the SPUs in the CELL figures, so it's a bit biased to compare a PPC +8 DSP-like accelerators against a single x86. The SPU are fine for doing some DSP acceleration, they are of little value for a desktop plarform used for text processing or surfing the web.

Reply Parent Bookmark Score: 2