Liste des Groupes | Revenir à col misc |
D <nospam@example.net> wrote:My thinking was more pixels to update, more loads on the computer, coupled with some power to handle the touch aspect as well. But you are probably right. As for screen size, my favourite form factor is 11.6" which does not exist any longer. =( Only 13.3 inch or 14 now for general use laptops.On Thu, 17 Jan 2025, Computer Nerd Kev wrote:>ARM SoCs for "ultra-low power" exist for things like smart watches,>
laptop makers just don't use them. Note that the big screen on a
laptop is a major power draw too, although a very niche product
using a large eInk display with some of the new partial-update and
colour features might be useful for some people.
Very interesting. Thank you for the information. As for eInk I haven't
(yet) seen one that would be able to keep up with videos. If such a thing
exists, it would be awesome!
No way with videos, but then I very rarely view them on a laptop
anyway, so I'm more interested in things like reading/writing to
Usenet. I gather the writing part has previously been shaky because
eInk displays update slower than people type, but new ones can
update one section of the display faster than refreshing the whole
display, so you might be able to practically update around the
cursor posiiton without such excessive lag.
>When I buy laptops I always try to get the lowest resolution I can, as>
well as to avoid touch screen, in order to increase the battery life.
I'd have thought that most power is used to generate the light
(from an OLED display or LCD backlight), so screen size would be
more related to power usage than resolution.
>Finding the minimum-power ARM chip able to run Linux would be an>
interesting exercise. I gather Linux requires a minimum set of ARM
extensions, as well as an MMU, which may exclude many of the SoC
options intended for embedded applications. Software is as much of
the problem as hardware.
I wonder if linux + modern work such as spreadsheets and audio/video calls
would be possible on those minimum-power ARM chips or if they are too
slow?
Well many are faster than the Pentium 1 PC I'm posting from now.
Clock speeds of 250MHz for example, while using a fraction of the
power from something like a Raspberry Pi Zero. This PC can do
spreadsheets and probably low-fi audio calls, but not video.
>
This is my thinking in saying the inability to use these
ultra-low-power SoC chips for more general computing is a software
problem. Unlike microcontrollers of old, these ARM-based things are
matching and exceeding the specs of this PC that I'm using. The
difference is that you can build most Linux software to run on a
Pentium 1 with one compiler argument, but it appears to be
difficult to build for these chips, and existing work is geared
towards running a custom, or at least customised, programs.
>
After thirty years minimum-power processors seem to have caught up
with the minimum level of processing speed that I regularly find
use for. Only by the fact that the architecture is wrong for
existing software does that preclude me taking advantage of this
advancement, because I don't want to try and rewrite tons of common
x86 software to run on these embedded architectures.
>
RAM is also a limitation, and since it's locked inside these SoCs
you can't expand it. Again projects like Contiki OS chew up vastly
less RAM than Linux while still offering networked applications and
a graphical environment (in the old version).
>
>
Les messages affichés proviennent d'usenet.