Display memory = 1400 bytes on the DS2000 = 1200 bytes on the DS1000Z= 50 pixels * screen divs * 2
So actually it's even worst, heck ... i'm embarrassed, we have also a ridiculously slow interface, no sh*t !
Sorry for the french, but on the base of what you report it would be enough to deal with a mere 72 KB/s data exchange to maintain a remote live screen 4 channels @30FPS (600*4*30) and as you guys have proved, the DS1054Z is actually a loooot slower!
I can understand that those physical interfaces are there mainly for instrument setting, but hell, these are numbers suitable for an Arduino UNO board.
The performance of the CY7C68013A USB controller that we can find in 60USD 20Mhz DSOs here seems an unbridled luxury (and they are transferring the whole acquisition buffer to PC).
Year ago i hacked one external PATA HDD enclosure based on CY7C68013A (that contains a sort of 8251 mcu) and attaching a TI 14bit 80MS/s ADC evaluation board on its original IDE PATA bus i achieved sustained 16 MS/s transfers (32MB/s) at easy.
Once properly programmed such controller requires only to write the 16bit word on its data bus and then put a transition on its bus clock pin (the controller code is upload from PC as soon as the device is attached, of course with proper device driver installed) so the external data source has not to be smart.
Ok, as far as i understood, no remote live screen is possible with mine DSO (DS1074Z), so i'm definitively on the market to buy another one.
I should have a sit in front of a real scope screen to evaluate, but right now i do not know how.