how using gpumemory for zram or ramdisk

i know its pearls for the pigs.
But in many servers are nvidiacards that they dont use it.
thats why here my question:
how can i use a piece of the gpumemory for zram or ramdisk?
the lzo-compression should also worked on the gpu.
So are this machines will be boosted i think and nvidia have the nose before amd.
im not very though to program this modules .
it would be nice to find a solution in the next SDK-examples.

thanky you,

today we have snow again here in Germany :-/

Video cards are meant to have a good inbound speed (cpu -> videocard), because you upload textures to it, the opposite is often not true.
In my tests i’ve got about 10MBps when reading from videoram.
I tested a 7900GT, a 9500GT and an ati 7200.
The fastest was the ati.

Another problem you will face when using gpu memory as system one, is that you could write in some areas that are currently used by the graphic driver, so you will end up with a corrupted screen.

Ati opensource driver allows you to limit the amount of memory used by the graphic, and with a few calculations you are able to avoid conflicts, but this is not possible with nvidia blob, so you’re out of luck anyway, maybe nouveau allows this, but the speed problem remains.

You may still want to read:
http://www.gentoo-wiki.info/TIP_Use_memory_on_video_card_as_swap
https://wiki.archlinux.org/index.php/Swap_on_video_ram