I wonder the
following thing : can we route the PCI channel through an other
channel ?
A converter or an adaptater like in an external hard-disk (IDE to USB),
thus, something like PCI to Ethernet.
By this way, the computer would have two Ethernet connectors, one for
the screen, one for the network and not necessarly a PCI slot inside.
(And if one can or has to plug the keyboard and mouse on the screen,
that will be good)
Is it possible ?

Computers can and often do have multiple Ethernets.  Yes, they make
PCI cards that supply Ethernet ports.  You can even get PCI cards
with multiple Ethernet ports.  I've seen PCI cards with 4 ports.
Some mainboards have more than one Ethernet onboard.  They also make
USB-to-Ethernet adapters, so you can get an Ethernet port without
using up a PCI slot, assuming the computer already has USB.


The computer would not have to add to an already existing ethernet port - the screen could be added as part of an existing ethernet network. Or, on this MacBook, I use Wireless 802.11g for internet, and can use the spare Ethernet as a high-bandwidth no-contention connection to the monitor - and it could easily play very high quality video.

You don't necessarily need a dedicated Ethernet to run this, even if
you are viewing HD video.  OTA HD mpeg2ts is approx 19 Mbps max.
That doesn't fit on 10 Mbps Ethernet, but is only about 20% of 100 Mbps
Ethernet, or 2% of Gigabit Ethernet.  This assumes that the decoding
happens at the terminal, not in the host computer.

I didn't know enough about the video specs, but numbers like this mean that a networked decoder is a very real possibility. Maybe even wireless with 802.11g, but ethernet is a much more consistent and secure standard, not to mention more widely used on desktops and laptops alike.


So a computer with only 1 Ethernet could connect to a hub/switch/router
type box and the terminal can connect to some hub/switch/router type box.
The terminal is just another node on the network,  The terminal can talk
to any computer on the network you like.  No KVM switch needed.

This is exactly what I was thinking. here is a response to luc's off-list response involved with this "networked viewer" idea:

The ethernet idea is an interesting point. One could simply attach  
the monitor on a Gigabit network, or on a custom, say, LVDS  
connection, and use that to send commands to the screen/graphics card  
assembly. 
A precision please, is it not like Video Network Computing ?

Yes, the idea is similar to VNC, except that you could provide easy 3D acceleration (VirtualGL is very similar - http://virtualgl.sourceforge.net/background.htm), and also support for high-frame rate video.
So, yes, it is very similar to that, except that each node is specialized to drawing. I was thinking along the lines of just sending kbd/mouse events to this "X server", and not receiving compressed images back, so the network activity is very low. Of course, the computer sending Keyboard and mouse events would also being doing all the processing, and send X client stuff as well, but it is a faster version of VirtualGL in theory.
The other approach is to just have this monitor act as a networked accelerated display. This could lead to video conversion on-the-fly, impressive 3D effects, etc.
taken from here:
a gigabit ethernet connection is just slow of being able to emulate a full speed PCI connection, while a 10gigabit ethernet can easily support AGP 4x and PCI-X, as well as PCI Express x4. 10 gigabit ethernet is rare and expensive, but considering how fast these ethernet connections are, tunneling things through them instead of a PCI connection should not be difficult.

Actually, now that I think about it, running X on the card is probably not worth it. Considering all the trouble it causes, and the fact that X is not nearly the bandwidth of things like 3D and high quality video, running those better is worth more to me than running X on the screen. The advantage of running X on your computer, of course, is that it follows you between screens.

That way, you could attach a bunch of monitors and machines  
to a network together, and since each graphics card is running the  
actual X server, it is simple to have tutorials, etc. In this ideal  
network, the "X server" is _actually_ a server, and the client is  
actually a client.
If LVDS is more of your thing, the computer could send commands over  
something as small as a phone cord or USB Cable (Well, you would  
probably use more control signals than that, but maybe Ethernet), and  
make it very easy to attach.
Ah! it´s very appealing indeed even if one needs the X-server one the
g-card (or elsewhere, anyway).
Nonetheless, the idea to route the PCI channel through an other channel
seems to be more exciting for me. A converter or an adaptater like in an
external hard-disk (IDE to USB)

In theory, one could implement other standards over a network connection: see the Keyspan adapter:
I agree that it is much nicer, both for the developer and the consumer, if the signal is routed through an already existing connection. The best options, due simply to bandwidth, are Ethernet, FireWire (800), and USB 2.0. FireWire is not on almost every computer, where as USB 2.0 and ethernet are. Ethernet, however, is most likely the better option due to aforementioned reasons: many clients for the single screen (it could seriously become a competitor to the Media Center, as in the concept and not the XP edition, if good audio support and TV in were included), super-high bandwidth (you could send, realistically, about 40 mpeg2ts signals over a gigabit connection), and integration into existing software (e.g. X). In fact, if video decompression was added to other OpenGL support, it would become a very nice multimedia tool that would be able to play the latest games, watch great TV, and also be useful as an easy-to-use computer monitor, something that no product currently in production is capable of (right?). (I have no connection to VirtualGL (sadly :-) ), or Keyspan)


nick

_______________________________________________
Open-graphics mailing list
[email protected]
http://lists.duskglow.com/mailman/listinfo/open-graphics
List service provided by Duskglow Consulting, LLC (www.duskglow.com)

Reply via email to