PS: sorry for the double post

IO-bandwidth has nothing to do with this.

What you need to display such a screen are two things:
A) Chip-mem bandwidth = This is no problem
B) Clockrate for the Video DAC

You are making AGA twice as good as it was. ;-)

AFAIK, AGA could NOT do 1280x512 in real 50 Hz.
AGA could only do this mode interlaced = which means 25 Hz.

1920 x 1080 @ 60 Hz in 24bit = 23 times the AGA bandwidth.

Cheers

I wonder how much faster SuperAGA will go than AGA?

Would SAGA 1920x1200 @60Hz in 8bit (multiplied gives 1 105 920 000)
need the following bandwidth increase on
AGA 1280x256 @50Hz in 8bit (multiplied gives 131 072 000)?
(1105920000 / 131072000) = 8.44x

Is that an accurate way to work it out or is it more complicated?
If it is okay is 8.5x in the realm of SuperAGA?

Could SAGA do the less colours (8bits) more resolution (1920x1200) tradeoff like OCS did?
ie. 640x512 16 colours, 320x256 32 colours.

Or is that too wasteful in itself?

25 hz is progressive video, 50 hz is interlaced (or 50fps progressive where available).

As I said before:
I have no fear that SAGA can reach 30 times the AGA bandwith.

But for huge resolutions you also need very high clock rates for the pixel clock. Getting to the needed pixel clock for HD1080 is challenge.  We will have to wait for the final board to see if this is feasable.

So.... IF someone wants to go extreme (spend $100 to $400 more) is there ANY component that could be upgraded on the motherboard to make a NatAmi hotter for demonstration purposes?

Overclocking NOT being an option.

I hope one day I can contribute something to the Amiga community.
At the moment I'm too busy working on my own projects.
It is for this reason that I as a fan am grateful to you and the team for their efforts at putting the train back on the track.
May it shine brightly!

I am surprised that you intend on releasing the next version next year. The developper version will maybe be out this summer, the date is not even fixed, then the developpers will need at least a year to produce software before the consumer version comes out. That year is also necessary to make sure all bugs are fixed, consumers will not wish to have to update the FPGA especially since there are less tech savvy than developpers. I see it as a real luck if the consumer board is ready for release on the Amiga 1000's 25th anniversary, July 24 2010, which would be a good marketing point, I have the impression that the release will be later than that.

After that, seeing the amount of work that went in the first release (if the consumer version is out in 2010, that will be seven years). I would be surprised if the next version takes less than at least half that time which would bring us to Q1 2014 for the next consumer version. And I have to say that I honestly hope that by that time the NatAmi project will be using integrated circuits rather than FPGAs.
 
> But for 1920x1024@60 Mhz you need a pixel clock of nearly 166 Mhz.
I recall reading that the official pixel clock for TVHD is 162mhz. Anyway, mabe it would be possible to solve this by putting an external pixel clock? Anyway I have the impression that simply by pushing the design to its extreme limit it may be possible to get that sort of resolution. Another idea would be to release a high end edition of the NatAmi in which, the chipset would be clocked faster and liquid cooling would be used to compensate.

Also, even if the NatAmi does not support FullHD, as soon as 1920x1080x30p/i is supported, it is "HDTV compatible", and being HDTV compatible does not seem a totally unrealistic goal to reach, this is only half the bandwith.

Finally, to reach HDTV compatibility, maybe an approach would be to write a player in which MPEG1 / MPEG2 / MPEG4 AVC / VC1 would be offloaded to the 3D core, which is basically a vectorial processor, same thing for motion compensation, and where the software would be hand coded in assembler for maximum performance.

This would require YUV support though, I am not sure the NatAmi supports YUV.

1920x1080 with 4 colors would be great for CED!

I'm not pushing for 1920*1080 capability to embarrass NatAmi's "slow" CPU, but because it's "the future" and the "the future" is essentially now.

In 2 years, almost everyone will have HDTV class TVs.

If it can't output 1920*1080, well, it can't, but if at all possible.... somebody will find a way to exploit it to some uses.

Natami is a computing platform, what's wrong with using a computer monitor? If you really must use a TV there are/still will be non-DVI connectors on HDTV class TVs. I can't work out if your issue is a lack of DVI port or whether you really have some urgent need for those extra pixels. The image quality from the Natami will be better than classic Amigas provide, there are cheaper solutions for watching HDTV content, what benefits to the computing side (not watching films) do the extra pixels provide to you? I'm interested to know.