Doug wrote:
>
> >[Updated Jan. 22, 2003] IBM unveiled an open standards based Linux/Java PDA
> >reference design at LinuxWorld in New York this week. Based on the IBM PowerPC 405LP embedded processor, ...
> >The e-LAP reference design PDA includes . . .
> >List of 17 items
> >. PowerPC 405LP Processor
> >. 32MB SDRAM
...
> >. Philips USB 1.1 (one client port, two host ports [one is disabled])
> >. Xilinx XCR3128XL FPGA (128 macrocells, 3000 gates)
...
> >IBM's PowerPC 405LP system-on-chip processor, which seems poised to compete
> >with Intel's XScale processors for mobile device design wins, is described
> >in
> >its preliminary data sheet as a "highly integrated device offering
> >high-performance at ultra-low power". The chip contains a 32-bit PowerPC
>
IBM's 405LP, due to several architectural enhancements, is the ONLY
PowerPC that can compete with ARM for hand-held devices.
Although current production is on a .18u fab, it is scheduled to move to
.13u technology which will make it economically viable to sell at the
target price ... which is significantly less than X-Scale.
Notable features include a voice-processing front-end for
voice-recognition that can be "switched off" when not in use to conserve
power. Also triple-DES in HW for secure communications, etc.
RE the funny USB note: "Philips USB 1.1 (one client port, two host ports
[one is disabled])"
That chip is designed for On-the-go (USB-OTG) and has one HOST and one
HOST/Client port, so it doesn't really have 1 client plus 2 hosts. You
have either 2 hosts or one port of each type as IBM has configured it.
Note also the bogus LCD controller specs:
"color LCD controller for 1/4 VGA up to XGA (2K x 2K pixels)"
Why do manufacturers state the resolution that the controller's control
registers will accept and not the resolutions that the chip can
support? Almost all LCD controllers are limited by memory bandwidth.
For example, Sharp's ARM-7 and ARM-9 CPUs have the same LCD controller,
but the ARM-7 can't support the same resolutions as the ARM-9 because
its slower memory bus won't fetch pixels fast enough. And the old
StrongARM CPUs didn't work properly above VGA resolution because of a
memory controller priority bug, although the chips are advertised as
having a high-resolution controller.
Final gripe: why put only 32-megs SDRAM on this thing? We're putting
64-megs of Mobile SDRAM on 90-MHz ARM-7 CPUs. This puppy should have 96
or 128 Megs of RAM to let it do what it's capable of!
Brian
~~~~~~~~~~~~~~~~~~~~~~~~
Brian Empey, P. Eng.
President
Technical Solutions Inc.
~~~~~~~~~~~~~~~~~~~~~~~~
--
Subscription/unsubscription/info requests: send e-mail with subject of
"subscribe", "unsubscribe", or "info" to 
Wear-Hard Mailing List Archive (searchable): http://wearables.blu.org
Please, *PLEASE* don't subscribe through a forward/expander/false domain
From Wear-Hard Mailing list Archive (WH)
Maintained by R. Paul McCarty
Archive created with babymail