qq1975b wrote:This will require an ATI Mach 64 ISA, right? very difficult to find one
Building a VME gfx card based on fpga cpu is lot more complicated, I suppose
It is more complicated. Nature did so for the CT60/63 with their SuperVIDEL. It's 100% done with some ram and a FPGA.
I can do basic CPLG stuff but designing a video card using a FPGA is not something I can do (may be one day).
If Nature was to make one the question is then how many people would buy one ?
I looked around at open hardware stuff and most of the FPGA video "card" design available out there are very limited in resolution.
So we would need to find someone that can design a video controller, probably with both VGA and HDMI/DVI output and a decent resolution (at least 800x600 in 65K color, ideally 1024x168 in 24 bits. All of this without knowing how many they can sale.. this could become costly for the person/company doing this. Add to that the software side as we would need drivers (I thing doing a FVDI driver is the best option there).
So this is a lot harder than just doing an ISA VME adapter for which we already have the schematics and find the right card for which we have the driver. If we want different cards we can then look for datasheet of video controllers and write FVDI drivers.
The lack of VMD graphic card has been an issue for a long time and a cheap solution to this would be more than welcome. The current VME-ISA schematics we have can be adapter to the Falcon bus easily as well as the Mega ST bus. For the STF, we can make a CPU adapter as all the signals are there (we would need to encode the interrupt signal but that's not hard to do). For the STE.. we would need a PLCC adapter (or a PLCC to DIL adapter and then use the same ISA adapter on it).