Re: Debian ARM success story: Debian desktop on a TS-7300
Bill Gatliff wrote:
> Lennart Sorensen wrote:
> ARM on the other hand who knows. I suspect it still stick
>> around in small embedded sytems along with MIPS and PPC.
> Rmk says he's got an SMP ARM machine, so maybe they're headed upwards
> OT: I'd love to see MIPS stick around, tho. Not saying it isn't, but...
ARM is the most successful semiconductor company on the planet and one
of the most profitable. ARM LTD. focus on fabless designs and licenses
there processors to any company for building micorProcessors on including
on substrate (such as ASICs or FPGA....).
ARM has whipped the snot out of the worlds leading 8/16 uP company,
destroying the margins in 8/16 bit processors Any start can get very
ARM technology and have their own processor. That's why Motorola (now
has dumped the 8/16 bit market and even now licenses processors from ARM.
Likewise TI licenses technology (ARM CORES) from ARM LTD. ARM is steadily
building up cores for larger and larger needs. ARM will survive and
they have the ultimate business strategy. Employ the best, Like Russell
father of Linux on ARM...... The technical folks at ARM are elite. Truly
of 'ass_whipping dudes'....... The accountants take the names.
MIPS has been floundering for quite some time.
MIPS is technology that is 99% in the Public domain. This means no major
is going to invest in a architecture that is not patentable. Back in the
Day when MIPS
was married to Silicon Graphics, they use to sue the same tactics as the
of the SVR4 linux licenses, Sue and intimidate. When SGI fell, MIPS was
to the laboratories and impoverished startups with core experince in
MIPS. You can build you
own MIPS CPU:
As has been stated PPC has a very bright future. Freescale has deep
with the other Motorola sister companies. Motorola split up not because
it wasn't profitable,
but because it allows the enormously wealthy family behind motorola to
'renig' on it's
finacial retirement oblgations to it's employees. PPC is the favorite
processor, by far, of the
US DOD. Lots of brain_dead US DOD engineers will use nothing else. PPC
nitches that other processors cannot fill, and in many circumstances
with the Freescale DSPs keep them in contracts for a long time into the
future. If you
look at the PPC offerings they are continually adding new, high end
commercial linux and 'wheelin&Dealin' in the financial markets with new,
technology centric to the PPC. The Mac computers, are dying off, with
the rise of
linux. Most of the MAC technical users I know at least dual boot into
acquired MACs so that there would be a third (pathetic) competitor to
Linux is whipping the snot out of MS in the server markets. Windows does not
even get mentioned in super_computing. And IBM runs linux on their
fot the DOD now.... Bill Gate is moving to philanthropy because he can
deals for philanthropic funds to dirty deals with winbloz in foreign
no saint, he's a scum_bag. His pal Warren Buffet has just about
purchased all of the
'dark fiber' in the US, at pennies on the dollar.
The prices dropping out of the bottom on x86 architectures is why Intel
is leaving x86.
They cannot compete anymore in the x86 arena. That does not mean that
x86 is dying,
many other companies have x86 based technology and cores that can be put
(FPGAs, asics, ....).
The real market for embedded is 8 and 16 bit processors that are in
wrist watches, alternators,
temperature sensors, etc etc. there about a 1000 to 1 ratio of 16 bit
and below processors
to the large well-known processors deployed world. This ratio is
increasing, in the favor
of the smaller uP every day. (note 32 bit processors will be consider
small, in a year or so).
Many uP have less that 2K(bytes) of code space. Those chips with state
do not have room for a RTOS, let alone embedded linux.
Don't get me wrong, I love embedded linux, but readers of this list
would benefit themselves
by going to web sites such as www.opencores.org and learning about
hardware, how it
is put together, and the resulting data sheets and example code that the
companies publish. You could learn how FAT file system work on various
chips which make up the various offerings from vendors. Spending some
time doing this
would allow one to develop expertise so you can separate hype from bullshit.
--in case anyone cares,