I'm Gregoire Gentil, the founder of Always Innovating and we have
unveiled earlier this week an ARM-based netbook:
http://www.alwaysinnovating.com
We got very interesting coverage on the web:
http://www.alwaysinnovating.com/touchbook/buzz.htm
and I wanted to share the praise with the community as the Touch Book is
heavily based on the Beagleboard. I have numerously stated my gratitude
to the work accomplished both by TI and the open source community
http://www.alwaysinnovating.com/company/thankyou.htm.
You may recognize the Beagleboard on the Gizmodo article photo:
http://i.gizmodo.com/5162584/always-innovating-touch-book-is-part
+netbook-part+tablet-open-source-frankenstein
the center of the board (which is our previous revision) is a
"duplication" of the Beagleboard. We have actually started our design
from rev. B5, we have added a 8.9'' LCD support as well as touchscreen,
accelerometer and USB hub. The device has 256MB too, so we are somewhat
equivalent to the coming rev. C.
Though it's very similar to the work accomplished by Gerald and the
hardware team, the schematics of our prototype is available here:
http://www.alwaysinnovating.com/company/opensource.htm
In terms of software, we are very close from Koen's Angstrom though we
will have a different front-end. We will release in the coming weeks
some interesting (I think) patches we are working on.
You should consider the Touch Book as a Beagleboard with a very nice
case - designer is Fred Bould. The plan is to start shipping in June and
we currently accept pre-orders for the US only. First batch will be
limited. I know that the Beagleboard community is spread all over the
world so we will accept international orders very soon. That's
definitely the plan.
I hope that the Touch Book will be beneficial to the Beagleboard
echosystem. If you are a Beagleboard hacker and would love that your
friends or relatives use a system based on OMAP3, I think that the Touch
Book will be a very good fit,
Gregoire
Following the previous post about the Touch Book
(http://www.alwaysinnovating.com), I would like to make an endorsement.
If you are looking for a hardware designer, especially for an OMAP-based
system, I stronly recommend Yuli Magniel who is an independent
consultant located in the Silicon Valley. Yuli has helped us to design
our Touch Book board and I must say that he has done a fantastic job.
His email address is in cc if you want to contact him,
Gregoire
PS: Additionnally, I would like also to thank Matt Evans, a UK-based
hobbyist, who made some miracle for a very early prototype we built last
year.
> On Mar 9, 7:26 pm, G2 <grego...@gentil.com> wrote:
>>
>> On Mar 9, 8:23 am, John Beetem <johnbee...@yahoo.com> wrote:
>>
>> > On Mar 8, 10:37 am, G2 <grego...@gentil.com> wrote:
>>
>> > > Note that you still have some options of the USBtoVGA dongle
>> > > (which is supported by the kernel) and starts to work very
>> > > well. But it's an additional cost for the user which is
>> > > unfortunate,
>> >
>> > Which USBtoVGA dongles are supported? Do you know which chips they
>> > use?
>>
>> Thomas Winischhofer did an amazing job when he developed back to 2005
>> "sisusb - usb kernel driver for SiS315(E) based USB2VGA dongles" which
>> is in the kernel. I did some tests a while ago and it can work very
>> nicely. To reply to Frans above, HD will be borderline and will
>> probably be very intense on the OMAP itself - though if you do your
>> video stuff in DSP, you starts almost at zero CPU on the cortex side
>> before doing the heavy USB task.
>>
>> I know that not all dongles work but I bought one at Amazon which
>> worked (the Tritton works I think),
>>
> Gregoire, when I mentioned HD, I actually meant 720p, not full HD. I
> know full HD (1080p) will be a very tough cookie, but 720p apparently
> is feasible.
720p is impossible over USB. Even at 16bpp and 24 fps you'd need
about 350 Mbps. USB simply doesn't go that fast.
--
Måns Rullgård
ma...@mansr.com
Without starting a great discussion on USB bandwidths, just a minor
clarification. The USB2.0 specification gives the possibility to transfer
data using bulk out endpoints at a rate up to ~57MB/s = ~456Mbps (as well
dependent on actual data content).
The overhead added by the USB2.0-protocol (electrical as well as SW) is
actually much lower than anticipated by most persons. If the IP-blocks - not
to mention the SW - can handle this transfer rate is a complete other topic.
But as such the USB2.0 standard itself allows the 350Mbps transfer rate with
no problem... :-)
I totally agree, that it will - from a system perspective - be very doubtful
if you can get an actual RAW data throughput rate of 350Mbps for most
systems - OMAP3 included...
Last example we have seen on this is the ongoing thread with respect to HDD
transfer speed, where top speed currently is around 6MB/s = 50Mbps, which is
really far from the actual USB limitation... :-)
Best regards - I hope this clarified more than it confused :-)
Søren
>> 720p is impossible over USB. Even at 16bpp and 24 fps you'd need
>> about 350 Mbps. USB simply doesn't go that fast.
>
> Without starting a great discussion on USB bandwidths, just a minor
> clarification. The USB2.0 specification gives the possibility to
> transfer data using bulk out endpoints at a rate up to ~57MB/s =
> ~456Mbps (as well dependent on actual data content).
>
> The overhead added by the USB2.0-protocol (electrical as well as SW)
> is actually much lower than anticipated by most persons. If the
> IP-blocks - not to mention the SW - can handle this transfer rate is
> a complete other topic. But as such the USB2.0 standard itself
> allows the 350Mbps transfer rate with no problem... :-)
I am aware of the theoretical capabilities. Nevertheless, I have
never seen anything manage to transfer more than about 250 Mbps over
USB2.0.
--
Måns Rullgård
ma...@mansr.com