some thoughts on SAKC

Carlos Camargo cicamargoba at gmail.com
Sat Feb 27 11:36:49 EST 2010


Hi

Wolfgang:
>
> > I have some problems seeing a big potential market for SAKC, so feedback
> from
> > someone like you is actually very important. If you cannot wrap your head
> > around SAKC, become excited about it and see some potential use in
> education,
> > then something is probably wrong with it :-)
>


SAKC is basically one NANONOTE without case, right now NANONOTE is a closed
platform, I mean, you can't interface Nano with the external world in easy
way. NANONOTE is a perfect tool for software development, and because of
that, we have a lot of Software developers in our lists, and just three
hardware developers. So, the question is: We want to attract hardware
developers? or the Qi target is just the SW side?

Projects like MilkyMist are very specialized and require a big background
and specific skills, rare to find. So I feel that Right now the HW side
needs recruit more HW developers. And SAKC will make this task. Without
hardware developers in our community projects like AVT2, SAKC, has no sense,
just sell NANONOTE as SW development tool. BEN NANONOTE is not useful for
derivated commercial products, you need a lot of interface signals/ports in
the real world. SAKC wants to fill this gap.


About the Wolfgang question about special connectors:

SAKC is a general board, we can design a lot of interfaces or "shields" for
different applications, of course, one of them will be one interface with
Analog outputs/inputs with special connectors. We will design daughter
boards for robotics, measurement, control, applications.




> Wolfgang:
>
> > For now the one thing I attach myself too is the low price. I think a
> good
> > quality board with CPU + FPGA for 99 USD is rare, if it exists at all.
>
>
Ingenic's CPUs don't have a lot of peripherals, just the most popular for
multimedia applications, If you have an specific Hardware task (FPU, PWM,
communication Unit, etc) you can't use this processor, so, you must buy a
commercial IC and interconnect it with the CPU, so, Again you need a lot of
interfaces. Another solution is design your own SoC platform, but Milkymist
work is in progress, and you need wait for it :). Another option is write
your periperal using HDL like Verilog or VHDL, sintethizing this code into a
FPGA and interconnect it with you processor. SAKC's FPGA allow the last
architecture and allow a lot of different applications that you can't
implement on ingenic's CPU.


> But then we have a huge software and marketing challenge, so I'm not
> sure...
>
>
You can find a lot of products similar to NANONOTE, a  lot of minilaptos,
Photo Frames, MP4 players based on the same chip, why you choice NANONOTE?
Because you cam modify the SW, you can build your own applications, and you
have the documentation for make it. But from HW point of view sakc is a
closed platform, you have the schematics, but if you want to make a little
change you need a lot of money and you must wait a lot of time for test the
new platform.

SAKC project is in the first phase, of course, you need a sw to run in it,
but, the software needed by HW developers is very different to the required
by sw developers, we can use NANONOTE, right now, NANONOTE is selled with a
basic kernel and file system. The same sw can be used on SAKC, u-boot and
linux kernel need little changes to run on SAKC, right now we've already run
openwrt file systems on SAKC pre-prototype [1], we run Ingenic's mplayer and
a lot of linux applications. We will provide the basic HAL for FPGA and user
applications for configuring it.


Concluding

1. A real HW and SW open project can't focus in SW side only.
2. Qi harware project need more HW developers.
3. SAKC's target market is different to NANONOTE's target market. Some
people can share both markets.
4. SAKC can be used for commercial products, people can midify it or
customize for specific applications.


Carlos

[1] http://downloads.qi-hardware.com/hardware/hacking/ECB_4725.jpg



> Bas:
>
> Yes, this is something that probably needs thinking.  I think there may
> be use cases for scientific applications, as I wrote, and in places
> where a standalone thinking unit is required, but a computer is too
> large and/or takes too much power.
>
> -----
> So much from Wolfgang and Bas about SAKC, work continues see
>
> Production schedule: http://en.qi-hardware.com/wiki/SAKC_run_1_schedule
> Software environment: http://en.qi-hardware.com/wiki/Development_Tools
> Status report:
> http://en.qi-hardware.com/pipermail/developer/2010-February/002120.html
>
> Wolfgang
>
> _______________________________________________
> Qi Developer Mailing List
> Mail to list (members only): developer at lists.qi-hardware.com
> Subscribe or Unsubscribe:
> http://en.qi-hardware.com/mailman/listinfo/developer
>



-- 
Carlos Iván Camargo Bareño
Profesor Asistente
Departamento de Ingeniería Eléctrica y Electrónica
Universidad Nacional de Colombia
cicamargoba at unal.edu.co
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.en.qi-hardware.com/pipermail/discussion/attachments/20100227/21c52330/attachment.htm>


More information about the discussion mailing list


interactive