[fedora-arm] 1ghz ARM Laptop (12in 1280x800 LCD)

[Gordan Bobic]

Luke Kenneth Casson Leighton wrote:
> On Mon, Jan 31, 2011 at 10:25 PM, Gordan Bobic <gordan at bobich.net> wrote:
> 
>>> with 1024x600 screens (the toshiba AC100, the Genesi-USA Ekiga and the
>>> AlwaysInnovating Touchbook) - i wish them every success in their niche
>>> markets that are catered for by 1024x600 screens.
>> Genesi Efika MX can take a 1280x720 screen.
> 
>  really?  hoooraaay!  wow, wow, and they cost about... $50 instead of
> $30.  i so don't understand why these aren't fitted as standard.
> still, bless 'em.  10in LCD, 800mhz CPU and 512mb RAM - it's...
> pushing your luck as a "main development machine"...

I, too, am puzzled by the choice of a low res screen. I suspect the 
reasons are:

1) $20 extra profit per machine is noticeable and/or reduced sales from 
a price tag of an extra $20 would be noticeable. The Efika MX Smartbook 
is already $350 which is quite expensive for a netbook.

2) Most people (the average users - I am not talking about the present 
audience of geeks who know what they are doing) simply expect a netbook 
to have a 9-10in panel that is 1024x600. Not out of technical merit, but 
because that's what all the ones you'll see at a shop will have.

As for "main development machine" - even 1366x768 would be pushing it 
for development. I wouldn't consider anything lower res than 1920x1080 
to be suitable for a dev machine nowdays.

My main work laptop is 2048x1536 (homebrew Thinkpad with a non-standard 
screen, google about it to get details on what to do and how if you're 
interested), but that panel is out of production and hard to get hold of 
at the moment, and it's fluorescent backlit, which means a >= 15W power 
budget just for the screen. :'( Then again, 15W is less than 1/2 of the 
power draw at idle when you have the rest of a Core2 system to keep 
powered up.

>>> for everyone else, who wants to see full documents and full web pages
>>> *without* having to press page-up, page-down, there literally is not a
>>> single ARM-based (or MIPS-based) product in existence, commercially
>>> available, anywhere in the world, despite a lot of talk from ARM, and
>>> also from the major ARM licensees, and despite the production cost of
>>> ARM-based and MIPS-based laptops being lower than that of an
>>> equivalent intel-based system.
>> You'd be forgiven for not noticing the difference. Genesi Efika MX and
>> Toshiba AC100 both cost me as much as a similarly specced Atom netbook. They
>> just have 2-4x the battery life and 1/2 the weight and thickness, no moving
>> parts (especially fans), etc.
> 
>  i know, damnit!!!  and the OLPC XO-1.75, which admittedly uses a
> 1.2ghz marvell "superscalar" CPU (which makes a biiig difference),
> reports are in that running Fedora it *outperforms* an intel atom
> laptop.

Yeah, the Marvell Armada (I bet Intel are kicking themselves for having 
sold their XScale ARM business to Marvell only a few years ago) is 
pretty good. My 1.2GHz Sheevaplug (Marvell Kirkwood) gives my Atom N450 
servers a serious run for their money at 1/4 the power budget.

>> So now I am pondering using a slightly newer yet very similar
>> W860CU. That is available with a 1920x1080 LED backlit panel (15.6in). It
>> wouldn't make much sense to use a higher drain screen when low power is a
>> key part of the appeal.
> 
>  yes.  that will mean that it probably requires a 3x LVDS IC [big, fat
> cable, at least 40 wires]  do you have a datasheet around?  unless it
> runs at 30hz refresh rate, in which case they *might* have only the
> one LVDS channel... yep, need a datasheet on the LCD panel.

I suspect (it's just a guess, but an educated one) the panel they use is 
an AUO (AU Optronics) B156HW01 or B156HW02. There are almost certainly 
several interchangeable variants of each, with possibly slightly 
different LVDS plug locations.

Here is a spec sheet for one of the variants:
www.yslcd.com.tw/docs/product/B156HW01%20V.3.pdf


[Clevo W860CU chassis: good]
[Tegra bad: bad]

>> The closest I have been able to get to finding a decent solution is the
>> Pandaboard (OMAP4 - Cortex A9 / PowerVR SGX). It has a PowerVR GPU which is
>> reasonably well understood, there is working XV video acceleration, and
>> decently working OpenGL ES drivers.
> 
>  _proprietary_ OpenGL drivers.

Which is better than _no_ OpenGL drivers (or OpenGL capability). Having 
said that, I don't actually care too much if there is working 3D 
acceleration, so OpenGL is definitely optional. But if there is OpenGL 
it should be the real thing, not ES. 2D acceleration with XV should be 
perfectly sufficient for the sort of thing we are discussing here.

>> With some luck we might get full OpenGL
>> drivers for it, too, since it is a GPU core virtually identical to what
>> Intel use (Intel GMA).
> 
>  ok - i've been researching the SGX hardware, and it is... complex.
> i've put in a proposal for funding, here:
> http://tree.celinuxforum.org/pipermail/celinux-dev/2011-January/002107.html

Excellent thread, thank you.

One thing I felt I should mention re: what is said in the linked thread 
- the Marvell ARMs aren't as good per watt as Cortex A9 based designs. I 
measured my AC100 (battery removed) at 6W at the wall plug, with no 
power management enabled. That includes a power brick (likely ~80% 
efficient), so let's say 5W. The Samsung panel in it is specced to draw 
about 3W, so call it an even 2W, for the Tegra 2 which is a dual core 
Cortex A9 with an nvidia GPU, RAM, etc.

My Sheevaplug (Marvell Kirkwood, no screen) draws 7W.

>  and i've been working to get SGX added to the GNU "High Priority" Project list:
> http://www.fsf.org/campaigns/priority-projects/index_html/#powervr

Good luck. My hardware reverse engineering experience is just about 100% 
non-existant, so I don't think I can help directly.

>  ok, unfortunately, there are two other things that stop the OMAP4
> from being used (and the Pandaboard).  the first is that you have to
> understand that 90% of TI's revenue comes from the military (radar,
> sonar) and this helps explain the rather top-heavy Free Software
> support as well as the high pricing.  the second is that the OMAP4's
> DSP speed and capability has hit some sort of threshold which has
> resulted in a BXPA "weapons" classification being slapped on it.
> thus, even to get samples shipped outside of the U.S. or Europe now
> requires permission - and a license - on an individual case-by-case
> basis, from the U.S. or U.K government.

Really? I had no idea. So how does the Pandaboard get around the issue?

>  the second thing is that the Pandaboard (and the Beagleboard-XM) use
> "Package-on-Package" RAM.  even TI's own documentation state that the
> yields on one memory supplier's 512mb POP RAM have been so rubbish
> that they got TWO percent success rate.  not to mention the insane
> cost of these high-density RAM chips.
> 
>  so, yes, you _could_ base a design around the (vanilla) BBxM or even
> the Pandaboard, but... yyeah :)

I don't see any downsides of basing it on the Pandaboard, except:
1) Low RAM amount
2) Lack of LVDS module for the next month or three

>  sorry, gordan, i really _have_ been on this for months :)  the
> S5PV210 is the lowest "from-scratch" development cost found so far
> (ok, not entirely from-scratch - it's actually adaptation of an
> existing, proven and well-understood 2009 CPU), and the next suitable
> one is the DM3730 / 3725.  which, because that's a 2010 CPU, is less
> well-understood so the costs are double.

I think supportability and availablilty should take more precedence than 
the development costs.

>  now i _have_ been advised of another two CPUs - one is the nusmart
> 2816 and the other is the ziilabs ZMS-08.  the nice thing about the
> ZMS-08 is that it is *already* available in a "system-on-module"
> format:
> http://www.ziilabs.com/products/platforms/zms08som.aspx
> 
>  use of this module would mean zero SO-DIMM development costs, meaning
> that all that would be required would be a motherboard, and that's
> only about $2k-$3k!

What is the CPU on that? Cortex could mean anything from Cortex M0 to 
Cortex A9.

And how would the RAM on that get expanded? How much RAM will the CPU 
support?

>  i'm investigating its price, availability and Free-Software
> compatibility.  it has a Cell Processor (8x8 vector processing unit)
> and ziilabs apparently are scared witless of being overwhelmed with
> whining eemo wannabe-developer support calls if they release the docs
> on its instruction set.  this is based on their experience,
> apparently, of commercial companies (asian factories) being completely
> incapable of programming it.  do the math on that logic, and you have
> to laugh, really, but i _do_ need to come up with a reassuring
> argument.

I don't see anything about a cell vector processor listed in the spec. 
If it does have it, then that may well be a BAD thing - more transistors 
means more watts, and the chances of any ARM Linux software using it any 
time soon is pretty close to 0 (look how much uses SSE properly on x86, 
and that's been around for over a decade, GCC still can't generate 
useful SSE code).

I'd stick with the _simplest_ possible Cortex A9 / PowerVR combo 
available. And I only say PowerVR because I'm not aware of more 
supportable alternatives at the moment.

>> Now, I recognize that 1920x1080 isn't what most people are expecting or
>> wanting out of an ARM laptop, and neither is a 15.6in size, but I like my
>> pixel density and pixel count. 13in would be better, but that panel is rare
>> and expensive and the only chassis that takes it is the Sony Vaio Z (at
>> £2500, not even worth considering).
>>
>> Now, obviously, even made using off-the-shelf-ish parts like this (you'll
>> have to work hard to get a Clevo chassis without a motherboard),
> 
>  yes :)  it would mean negotiating with clevo.

Or paying a small-ish premium for getting a bare-bones system from a 
reseller, and throwing the mobo away. If you add up what the components 
cost, the cost of the bare chassis with a mobo is tiny.

>> but the costs would likely be in the 3-4x what you were hoping for.
>>
>> Pandaboard is $174.
>> A Clevo W860CU is not far from 10x that (OK, that includes the motherboard,
>> CPU, an expensive GPU, RAM, disk, etc, but it'd be difficult to persuade a
>> distributor to sell you a bare chassis with a battery and screen).
>>
>> The upshot, however, is that you would likely get 20+ hours of battery life
>> out of it.
> 
>  ahh, it depends on the power

Sure it does, but we are talking about _maybe_ 5W for the screen and 
considerably less than that for the rest with a decent choice of 
CPU/mobo. I'd guess the total would be somewhere around 1/3 of what an 
Atom N450 netbook requires, and that is as good as it gets for x86.

>> The other thing I consider to be a big problem is the amount of RAM
>> available. Pandaboard comes with 1GB of RAM, which is on the small size if
>> we're really serious about this. I'd like to see at least 2GB, but I'm not
>> sure if this is viable with a Pandaboard. I haven't seen any ARM boards with
>>> 1GB of RAM:
>> Genesi Efika MX: 512MB
>> Toshiba AC100: 512MB
>> Sheevaplug: 512MB
>>
>> Pandaboard: 1GB
> 
>  POP - this will be eeexpensiiive.

$175/board doesn't sound that expensive in the grand scheme of things, 
unless you are referring to other potential issues (e.g. import/export 
licencing you mentioned).

>> Compulab offerings top out at 1GB
> 
>  ooo goood.

Indeed, but Compulab goods are _expensive_. 2-3x what you'd pay for a 
Pandaboard.

This may also be of interest:
http://www.reghardware.com/2011/01/26/compulab_trim_slice/

Still only 1GB of RAM, though. I wouldn't want to aim for less than 2GB 
to start with at the very least. And unless there is an _overwhelming_ 
cost or power consumption argument, the RAM should really be on a DIMM 
(it's hard to see how that wouldn't be the cheapest solution in the 
long-run).

>> Ideally what I want is something like a Pandaboard with a DDR3 SODIMM.
> 
>  *sigh* yehhhs... but then that means that the motherboard itself
> needs to be a 6-layer or even a 10-layer board, as the CPU itself
> needs to be on the motherboard.  and _that_ means you're now into $40k
> development costs (of months).  if the CPU itself is on the SO-DIMM
> (System-on-Module) and you can find one that suits already, then the
> dev costs are $2-$3k (and only a couple of weeks).  that's a _big_
> difference.

Agreed, but I don't think it makes sense to even consider this if the 
SoC module doesn't come with at least 2GB of RAM and Cortex A9 class CPU.

Just out of interest (forgive me if the question is daft, I'm not all 
that familiar how this segment of the market works) - is there a 
standard for the way SODIMMs SoC modules are wired up? i.e. would you 
then be able to replace the SODIMM on this custom mobo with another 
SODIMM SoC and expect it to "just work" provided all the features are 
present? I'm guessing the chances of this are less than 0.

>> But if something like this were available off the shelf, and based on
>> something very well supported by the community already, I'd happily pay a
>> considerable premium for it.
> 
>  ok.  the plan is to create at least a "generic" motherboard, approx
> 7cm x 9cm, with flying leads that can go to the connectors, regardless
> of how large the case.

Indeed, this is what I was thinking about with a Pandaboard - have USB 
(and audio) extension cables going to the case ports if needs be.

>  that way it would be kiiinda possible to fulfil other laptop specs,
> but the kick-in-the-teeth there is the LVDS and LCD power
> requirements.  the connectors vary *radically*.  i know of an IC that
> can do up to 30 volts (funnily enough the one used by the ODroid from
> hardkernel.com) but there do exist LCD screens, esp. the larger ones,
> that require more for their LED backlights (36v).
> 
>  it's tricky! :)

A bit of custom circuitry isn't the end of the world if you are shooting 
for a relatively stationary target. Having said that, it's only a matter 
of time before laptops go from LED backlit to outright OLED like the 
current phones, which may require a slight change to the circuitry. This 
_could_ be pre-empted by making the relevant components modular, of 
course. But then we could also say the same thing about the SoC module, 
and the chances are that the new/replacement SoC module would not be 
compatible.

>  but, yeah, the aim is to fulfil the largest number of peoples' needs
> first (weight of numbers) and then branch out from there.

I'm all for a project like this, but I suspect the volume will be quite 
low. That means high unit cost, and in that case it makes more sense to 
aim for the high end, since there is a higher chance of being 
competitive there. It will be easier to come up with a very low power 
and adequate performance 15.6in 1920x1080 laptop with 20 hours battery 
life for $1500 than a slightly better 10in 1280x720 laptop that has to 
be competitive against the $350 offerings.

Having said all that, it may be worthwhile having a word with Genesi. 
They are probably already working on the next gen of Efika MX. If you 
can stir up enough interest in something like this (and I would 
definitely be interested in something like what I've described unless 
the cost per unit ends up being eyewateringly outrageous), then it would 
make a some sense to try to get an established manufacturer on-board. 
Unless you have a very commercial/competitive venture in mind.

Gordan

P.S.

I think we should take this off the list, unless a lot of others here 
want to partake, since by my reckoning the Fedora related content in 
this is becoming pretty close to 0.