Ideas: The new Amiga. Yes, it's inevitable!

[Beast96GT]

I posted this in a previous thread, but it got no reaction... and dammit, I want a reaction!  

Here are just a few thoughts/ideas if the Amiga WERE to be redesigned, re-introduced hardware and software-wise:

1) The Amiga should be a graphical powerhouse.  It's an old idea, but it's amazing how much attention is paid to the machine with the best graphics--Hell, it's how the Amiga really made a name for itself (among other things, of course).  

Modern graphics cards are designed to be optimal for typical 32-bit floating point (matrices and such) math.  

The Amiga graphics chips (preferably on a card) could optimize for real-time ray-tracing.  It's not a fantasy by any means;  it's already being done on the PC, but it's limited and requires massive hardware.  The Amiga, if designed with the bottlenecks of doing this in mind, could possibly bring this market and exploit it as its own.  

The Amiga graphics chips could optimize for Quaternions which are much better than matrices (especially for rotations).  

It's amazing what graphics will do to get your platform some attention.  Remember the first time you saw the Amiga screenshots of Defender of the Crown?

2) Support your software developers. A good machine has to be supported by good software.  Good software has to be made with by developers and publishers who think they can make a profit.  On a new platform, help from the manufacturer is a must.

Reusable code is in that equation.  It's time to move past assembly code and support a higher level language like C++ with libraries.  This means a real compiler that supports serious debugging and a optimized machine language generation. Professional code-bases are huge, and require easier to read code.  

3) Have a multimedia solution like no other.  The Microsoft Windows PC has come a long way with multimedia integration, but in my mind it has never taken the next step nor done it right.  The Amiga could have additional inputs for things like HDMI and optical audio.  

The Amiga could have DVR and video editing capabilities.  Make people who own a video camera feel they need an Amiga to make (semi-)professional video of their ugly kids.  
 ;-)

4) Provide a modular approach. Being able to upgrade your computer is a good.  Computer manufacturers aren't always keen on this idea--they'd rather you buy a new system.  At least on your higher end system, it's a good thing.  You need to not be limited by your motherboard with what you can do with your system--especially graphics.

Like I stated in a previous post, this is really a pipe dream, but it's fun to ponder what would be if the Amiga has a serious investment.  
 :-D

[Beast96GT]

alexh wrote:
It would need to have integrated perpetual motion feedback joypad with a sub-cold fusion reactor for power.

Wow, might as well require a flux-capacitor, huh?

I disagree.

Real-time ray-tracing is already achievable with the right hardware on the PC.  A lot of the work can be done with CPU/GPU concurrency.  

My $150 TV has an HDMI input--why can't my computer?  Oh, that's because this computer is a PC and you need to buy a card for that.  

Why can't my $500 Geforce 8800GTX card do anything for me other than run games fast?  

Ok, yeah, I'm not a hardware engineer, but you guys are acting like I'm asking for the damn space shuttle!  

 :-D

Maybe that's why you're all still on here being bitter and complaining?   :-D

[Beast96GT]

Dave, those are some practical ideas, but in my opinion, nothing that will ever do anything for the Amiga.  Let's work on that list together, ok, bro?    

cicero790:  You've got an interesting question, but I would have no way of knowing.  Depending on how the new hardware was developed, this would dictate the performance of any game ports.  

[Beast96GT]

dammy wrote:
I find these threads rather amusing as if people have not gotten a clue that a new hardware is never going to be released that can make any tech dent in the current marketplace.  I find it even more amusing that people could even think the Amiga market is economically viable for any corporation to invest money in to make that dent.

Dammy

I think it's amusing how people like you feel they must contribute to the forums, yet you have nothing to contribute.  This thread is meant only for fun and conjecture--you may leave at any time.   :-)

[Beast96GT]

..double post..

[Beast96GT]

 

mdwh2 wrote:
(Reposting from the other thread, in case you missed it...)

Beast96GT wrote:
The Amiga graphics chips could optimize for Quaternions which are much better than matrices (especially for rotations).  

In what sense "optimised", that isn't already done on modern hardware? Presumably if there was some obvious way to speed things up here, NVIDIA etc would already be doing it.

Quaternions are only a representation of 3x3 matrices for rotations, they still need to be converted to 4x4 matrices to represent general transformations anyway.

Zekaric wrote:
Quaternions is just another tool in the box, it won't replace euclidean transformations; so they aren't exactly necessary IMO.

Well, if they're needed, they're written in software anyway, so they shouldn't have anything to do with hardware choices. Quaternions are better than Euler rotations as they avoid "grimbal lock", but use less variables than a 3x3 matrix (4 verus 9). It's also easier to do some algorithms using quaternions rather than rotation matrices (e.g., spherical linear interpolation).

To avoid gimbal lock and get smoother rotations, most game engines convert a rotation matrix to a quaternion, rotate it, then convert it back to a matrix.  You can't tell me this process can't be optimized, especially on the GPU.  

Chris