Oh! Thats inviting an reaction from you, so nevermind me.

Imagine being able to play games that looked and played AS AWESOME as those AND in higher resolution on a Vampire!!!!!!!

Give it time and it will happen. I don't see what the issue is. It will run the old software plus new software can be developed for it which will be more efficient since it will use its full potential, it's just up to developers. Once the Hard FPU is out for V4 then people can start developing for it. Progress will happen but it takes time.

Not impossible, or not the goal of the Apollo Team?

IF you want to stick to legacy design, there is Jens/Individual Computers or Mist (You already got Mist if Im not mistaken).
Apollo Team got their own take on it.

There are choices, and its up to the customer to decide its what they want.
The team provide us with hardware with increased performance/features compared to what we've had available before.
Its not really their job to provide us with xyz software.

Lets say you are correct in assuming there wont be any software developed for Vampire spesifically in the future.
The Vampire will still present us users with a cheap (relativly), easy to use new hardware that lets us use modern monitors and resolution with ease, good mp3 decoding etc.
Much better than my current A1200 with Blizzard 030.

What is it really that torments you with Apollo? The lack of MMU to the extent it wont run your favorite Linux distro?

Then go an develop that FPU yourself and don't expect others to make two FPUs, one suiting your preferences and one meeting the Team's goals.

The job you are doing for the V4 is really cool. But i've a question for the V2+. What can we expect of the partial hard FPU with the software FPU compared to a real 68060 FPU with old software.

For example, if i want to start an AGA demo that need a fpu on my 500 V2+ (when the AGA will be available) what can i expect compared to my 1260/95Mhz ?

I doesn't expect to have a FPU powerfully as a 68060/500Mhz on the V2 but if it can do the work as good as a 68060/100 with actual softwares this will be a total awesome thing for me.

The FPU needs a lot space in the FPGA.
The Vamp4 FPGA is twice as big as the Vamp2 FPGA.

Therefore fitting the complete FPU fits easily in the Vamp4.

In the Vamp2 you can always rely on FEMU.
The latest FEMU woks hybrid it on uses 68080 special hardware features and does part of the operation in Software (ROM).

Have you seen the benchmarks of latest FEMU?
FEMU performance several times higher than any 68882 FPU.

Regarding your demo question.
Its very hard to answer this exact as the end performance is always a combination of several factors.
- fast memory speed
- cache size
- gfx memory access speed
- integer performance
- floating point performance

If you compare a Vamp2 with 68060 than the Vamp2 has the advantage in 4 of these 5 areas.
+ Fast memory is faster
+ cache is bigger
+ gfx memory speed is higher
+ integer performance is higher
- Floating point performance is lower if using FEMU
  in the V4 the FPU speed is even or higher to a high end 68060.

No because FPU is not public yet, it will be once v4 vamps go on sale.

Just seen it on the V2 FEMU topic. Sorry for this outtopic and thanks for your answer. :)

But dont forget benchmarks is only useful if you understand what it is testing for normal software written.

One would test graphics to know if your computer is capable of playing a game for example.
I always wanted a GPU rating system for games.

For me compile time is also important. So test that over time.

MKV playback or encode 1080p using FPU might help.

For example some people use unRAID to make PLEX media server & Itunes server with HD Homerun tuner device so they can watch their own shows.

This is real world example. Need low watt usage.(24/7) But CPU powerful enough to transcode PLEX for playback smoothly. On x64 Passmark ~5000.

Plex media player source EXTERNAL LINK 

Video codecs are optimally done in INTEGER, and AMMX is designed to speed this up tremendously.

Clearly if a feature wasn't prevalent 25+ years ago it's not going to retroactively manifest itself.

The Vampire is going far beyond preserving compatibility and moving into the realm of the next generation. Taking advantage of new capabilities is obviously going to require new software and that presents a very welcome challenge.

Actually, that's not correct.  We now have a CPU that can execute multiple instructions per cycle.  This wasn't the case 25 years ago. So most existing Amiga software will run much faster.  The increase in clock speed paired with the Apollo core and its optimizations will make most classic software applications fly.  If that isn't a "help" then I don't know what is.

Not at all, as Gunnar has pointed out hundreds of times, the MMU is right there, it is modern and feature rich. As will the FPU be. Brilliant stuff. What torments me, and should worry other people as well, is not that I will not be able to use Vampire cards for my Gentoo/m68k cruft, but _why_ there will not be any Linux for any system running Apollo Core.

What makes the Apollo Core different from just about _any_ CPU design on the planet?