@Fransexy
Quote:

Man, the prices are orientative, do not take so serious.that i want to say is that people pay more for custom that for standar. Second: AGA is outdated compared with ati and nvidia chips but even aga has features that this chips not have.It´s the key add features that this chips lack, if you want to compete in the race of MHZ then sure you bit the dust.But we better than nobody knows that MHZ is not all in computers

Step away from the computer. Take a deep breath. Repeat several times until reality sets in.
There is _nothing_ AGA is advanced in today. Screens were a convenience that has since shown no real need for regardless of the sentimental urges. software based virtual desktops work just fine. There are currently no 'key add' features in AGA. Modern video cards are already far past the old Amiga chipsets.

You say 'MHz is not all' but then look at the grief taken over the uA1s specs for the price...but now you want something yet more $$, with unidentified 'features,' that would surely doom the A1 series due to price to anyone but the most extreme hardcore Amiga guys/girls. Not a great way to keep a platform alive. Good way to kill it entirely though.

A dual core CPU would be nice, as would PCI Express. But the OS would need to support both first, and AFAIK, AOS doesn't support SMP right now, so any 'additional CPUs' is a waste currently. You, nor any 'small company,' are unlikely to beat ATI and nvidea with a modern GPU, so forget that idea entirely, please.

The computing landscape has changed dramatically- custom chips were made _because_ no equivalent designs existed at the time, in anything close. That is no longer the case, and some people need to either acknowledge the fact that:
1. They aren't old enough to have jobs yet, so have no concept of what sort of $ is involved, or what it means to invest that sort of money.
2. Wanting something entirely embiguous without any justification, knowledge in the subject matter being asked for....does no one any good.
3. In the future, if you want something, defining what 'something' is is a bit more helpful than 'I want something new.'

I've also got to ask- what exactly is it that you don't get right now that you NEED from your uA1?


@Atheist RE: IBM and motoroloa 'screwing up our MBs'
Uhhh- no. Neither one made the Articia chipset. Without Motorola/IBM, if Hyperions logic about not 'competing' on x86 platforms held true, instead there would be no uA1, or it would be based on a still slow (comparatively) platform like StrongARM or other embedded (but slow by todays standards) CPU. Please think beofre placing 'blame.'
MAI would be a much better candidate for blame, or EyeTech of rmarketing the boards as 'complete.' (which has already been shown in other threads). For the itme being, be happy with what you have.

_________________
Are we not done with the same silly arguments and flames yet??!

@-Sam-
Quote:

Quote: Instant amiga from power on. That is all that's needed to give to old amiga feel back. Atleast for me. Even with a RAD disk you still had to spend time loading. You must have had one heck of a fast hard drive!

Perhaps he's talking about kickstart.

Would a flashcard boot OS4 significantly faster than a HD?
If so, it might be worth making it standard. It might at least give us something to brag about if we can boot into an Amiga desktop in 5 secs. And, no, I never ever leave my computers running, although I don't mind hibernation. Hey, what about hibernation? Is hibernation even possible on an A1?

Ok. Going back to my happy place now.

_________________
Leo Nigro, CTO Commodore USA, LLC
Opinions expressed are my own and not those of C= USA.
Commodore/AMIGA "Beautiful, High-Performance, Home Computers for Creativity and Entertainment."

@Atheist

Posts 14 & 17, eh? Well, heres a response to #14...

Quote:

Atheist wrote: I wasn't necessarily meaning a custom video and sound chip (nVidia, ATI, and Creative Labs would be impossible to beat), but a re-integration of "chip" ram, and an extra CPU that ONLY does that blitter, or/and whatever it was that made screen dragging possible. If something on top of that is possible, well, great.

Firstly, nVidia and ATI graphics chipsets both have blitters onboard, this is what is called "2D acceleration". I've been an OS4 beta tester long enough to have witnessed the actual update that implemented blitter support in the radeon driver. It was a marked improvement in general speed.

Secondly, the "screen dragging" trick was possible due to a number of factors. The first is that the standard Amiga modes all use the same pixel clock rate with different dividers. The result is that the lowres (320 width), hires (640 width), and super hires (1280 width) modes all have the same horizontal and vertical refresh rates. Therefore, mixing several modes in one frame is simply a matter of changing pixel clock divider in the middle of a frame. You still end up with a perfectly valid PAL or NTSC signal.

The second factor is the "copper". This is a co-processor that could be programmed to wait for specific lines of the display and move data into other chip registers, independantly from the processor. This meant that you could change the screenmode and palette in the middle of the frame without any work from the CPU.

Modern graphics hardware can have completely different pixel clocks for different modes. In fact, this is necessary to display the various (S)VGA modes at optimum refresh rates on most monitors. They also simply do not have this "copper" functionality.

You simply cannot mix the different clock-speed modes in one frame. It just wouldn't work.

Implimenting copper functionality simply makes no sense in a modern computer. It would be tied to the graphics chipset very closely, so if you made another "Amiga" in the future with ATI's latest chip, you'd probably have to make a whole new copper chip every time. But then again, the copper wasn't just used to write data to the graphics chipset, and therefore people would write copperlists that wrote data to specific register addresses, which just makes for more legacy baggage when you tried to upgrade the "new Amiga" in future.

You *could* emulate screen dragging in software. The CPU would have to convert different depths and resolutionsto the frontmost screens resolution. Highcolour screens behind 8-bit ones would look fugly, though. You would either have to scale the smaller screens up, or just show them smaller. It would use up more video memory (unless you wanted to do it on the fly, which really would steal a lot of CPU time).

If and when the Amiga gets a new 3D GPU powered graphics system, re-implimenting screen dragging will probably be a LOT easier, so there is hope. However, making a "custom" Amiga is simply not a valid or realistic way to achieve that goal.

Last edited by xeron on 07-Feb-2005 at 04:16 PM.

_________________
Playstation Network ID: xeron6

@Chunder

Preferably with both a lot of zeroes in front of the decimal point AND a non-zero digit or two in front of the zeroes

_________________
This weeks pet peeve:
Using "voltage" instead of "potential", which leads to inventing new words like "amperage" instead of "current" (I, measured in A) or possible "charge" (amperehours, Ah or Coulomb, C). Sometimes I don't even know what people mean.

@Atheist

& post #17...

Quote:

So, Agnus isn't really a graphics chip. It's only one chip, that isn't nearly as sophisticated as what's being done today.

Correct.

Quote:

Is it really that expensive to get a company to make these, that could be interfaced with todays GPUs?

It is probably impossible, regardless of cost. Agnus can work the way it does because it is designed to be part of the Amigas chipset. You simply cannot take an ATI graphics chip and make an "Agnus" that would interface to it. The ATI chip simply isn't designed to have this functionality "tacked-on" to it. Even if it was possible, it would be a hack, and like I mentioned before would be tied to a specific chip. Not a range of chips, but a specific chip.

Even if you COULD make an "Agnus" that interfaced to, for example, the ATI RV200 chip, what would be the point? Lets look at the features Agnus provides:

1) Controlling around 25 DMA channels

Well, the AmigaOne boards have DMA already. DMA to RAM, PCI DMA, etc. etc. No need for Agnus there. (Although some say we need April


2) Generation of various system clocks

The PPC already has clock functions on board. AFAIK they are of a higher granularity than the Agnus clocks anyway (could be wrong there).


3) Addressing chip ram

"Chip RAM" would presumably be the RAM on the graphics card, and any audio buffer on the sound card. They already have controller hardware built in. Not really sure what Agnus would do here


4) The copper

The copper has two basic functions. To wait for a scanline, and to move data into registers. Considering that screenmodes have different pixel clocks, the ability to change them in the middle of the frame is useless. Sure, you could change the palette of palette mapped screens, but why not just use a highcolour mode? Also, having copperlists means that the data registers written to by the copper have to be in the same place on all hardware that the copperlist needs to work on. Modern chips don't have any registers that map in at $DFF000, so old copperlists won't even work on it. New copperlists won't work on old Amigas. Different graphics cards and sound cards have different registers. The copper really needs to be part of a specific chipset like it was on the classic Amiga to be of any use at all.

I'll say it again; once AmigaOS gets a proper graphics API using the 3D GPU, with a wrapper for legacy apps, all these arguments will become moot. The hardware is NOT the problem.

_________________
Playstation Network ID: xeron6

@xeron

Interesting posts - one thought, though... how fast would a copper chip need to operate at? Could it be a software-controlled chip? (is FPGA the right acronym? One of the ones where you can dynamically change what it does, anyway).

@olegil

Well, I kinda hoped that went without saying

@thread

In effect, what would be required is a private investor/company/consortium/whatever to fund the development of something that is better than anything else on the market - because that is what made the Amiga special. And as others have pointed out, due to the money involved in the market already, doing a one-up on all those rich companies just isn't going to happen!

... unless you find some amazing new technology stuck inside you after you next get kidnapped and probed by aliens...

_________________

@Fransexy
>You are making a great work with OS4 but you have the same actitude as this people that not
>believe in the dreams only in the $$$$$$$.


You don't seem to have any idea how much this whole OS4 project compares to charity work... How many of the devs are actually going to see income surpass their expenses? I don't expect to. I just wanted a better graphics card than Voodoo3, anything else is "bonus". I wil be very suprised if I see enough income from this to pay for the hardware I've personally purchased for this project. If we were doing this for the sole purpose of turning a profit, we would NOT be doing this at all.

_________________
All glory to the Hypnotoad!

@Chunder

Quote:

how fast would a copper chip need to operate at?

Does it matter? The fact is that even if you somehow overcame the technical problems and made one still it wouldn't give you screen dragging, so whats the point?

_________________
Playstation Network ID: xeron6

(Responding to post #1)

@Atheist

I'm not sure what your questions...
>Is it really not going to happen again?
>Could it really be that expensive to do again?
...have to do with Viva and programming FPGAs and I'm surprised the thread completely missed the most interesting part of that post: FPGAs!

FPGAs may be slower than most current CPUs but they can be "wired" as needed to do massive parallel computing that would otherwise require dozens of CPUs and insane cooling.
Designing the layout is a huge task. It might have become a bit easier now with that Viva tool though, it looks good on paper anyway.

IMHO their true potential has yet to be realized (e.g. replacing the CPU and custom chips of desktop computers).
The transmeta partially does that and the C-One uses an FPGA to simulate some of the original C64 hardware, but the real fun will only start when there is an OS capable of reprogramming those chips on the fly.
:drool:

_________________
AlexC's free OS4 software collection

AmigaOne XE/X1000/X5000/UAE-PPC OS4 laptop/X-10 Home Automation

@Chunder

hardware C2P, hardware Coper to truecolor/fastram to speed up UAE emulation,
just put this stuff on PCI card like it's done on the "Catweasel MK4",

_________________
http://lifeofliveforit.blogspot.no/
Facebook::LiveForIt Software for AmigaOS

@NutsAboutAmiga

Quote:

hardware C2P

What for? Modern chipsets don't display in planar anyway.


Quote:

hardware Coper to truecolor/fastram to speed up UAE emulation

I don't think that would actually work. If you're that bothered about "hardware emulation" for legacy apps, why not come up with some sort of AGA compatible FPGA that can DMA into an overlay window on workbench? That would be kind of cool.. but it simply would not be worth the money or the effort that would be better spent on improving the OS. UAE is getting better all the time anyway...

I guess the problem is that people are pining for the days when the Amiga had superior hardware, but all the functionality they then go on to ask for is either possible but requires future OS updates ("I want screen dragging!"), possible with future updates but not really worth it (arguably also "I want screen dragging!" ), or just a result of them not understanding any of the technical aspects of what they're asking for ("I want a copper!").

Last edited by xeron on 07-Feb-2005 at 05:36 PM.
Last edited by xeron on 07-Feb-2005 at 05:34 PM.

_________________
Playstation Network ID: xeron6

@Atheist

In my opinion I would rather see the AA and other custom chip stuff put on a card for our current Amiga platforms. Even so, I am not really interested in running legasy stuff as much as seeing new applications for the platform. A JIT version of UAE on the system will probably make a project like that not even worth the work that would go into it. UAE is a better option. It would be nice to see it run a bit more seamless as well.

It's very simple everyone...in 1985 we didn't have highly competitive companies that annually invested millions of dollars in research and development devoted specifically to the development of only one aspect of a computer (ie graphics card, sound card, Northbridge, CPU, etc...). We didn't have one, much less several (ATI and Nvidia and Matrox and whoever else happens to be around in graphics for example) who devoted the majority of their energies to simply developing this one piece of a machine.

If we built a "custom" machine (whatever the hell that means) that even managed to be on par with the development plans of Nvidia and ATI (which extend as far as 6 years into the future), our technology would be quickly and easily superceded almost immediately as they impletmented their own variations of whatever "new" technology that was any good that we had introduced, and then continued along their EXTREMELY high paced and extraordinarily well-funded development cycle...then poof, our "custom" machine is once again under par.

When there are companies that devote their entire existence to developing, funding, and progressing each individual element of a computer, why would you even bother trying to take ALL of them on just to reinvent the wheel? A computer is best customized by its Operating System and available features...one of those features should be the ability to continually add or support new technology without having to buy an entire new computer...this is what busses are for, and I for one appreciate being able to buy a computer that has the potential to work with the latest and greatest cards out there because, once again...the companies that devote their entire existence to the development of one kind of card are going to continually have newer and better things to offer.

the underlying hardware in all PC's today is pretty similar, yet the experience of MacOS, AmigaOS, DOS, Windows, Linux, OS/2, whatever OS you prefer... is what makes your machine feel and perform differently.

Anything else you insist on simply demonstrates your lack of understanding regarding computer hardware, development, and the industry...and if you don't know anything about computers then you really have no business bitching about the way one of them is made.
Especially when the only hardware you apparently have any familiarity with is 20 years old, and shockingly dated and underpowered when compared to the combined forces of M-Audio, Nvidia, Creative Labs, ATI, Intel, AMD, Motorola, VIA, and anybody else who is devoted to the advancement of computer hardware.

This is a very tired and stupid topic...can we please move onto an intelligent discussion regarding Amiga's future.

Oh yes... everything Wain said, too.

_________________
Playstation Network ID: xeron6

@xeron
>why not come up with some sort of AGA compatible FPGA that can DMA into an overlay window
>on workbench?

That's what the BoXeR is. And we all know what a successful product this is.

_________________
All glory to the Hypnotoad!

If you want a custom computer, build it yourself

I'd like to see AmigaOS evolve into something more or less hardware-agnostic (?) and portable. There would be no sense in developing a "custom" hardware solution, at least not ATM.

Listen to the Hyperion guys, they know what they're talking about! It might be hard to grasp, but the people working on a particular item tend to know more about it than others

_________________
UNIX is user friendly. It's just selective about who its friends are.

@Rogue

What I think people mean when they say "custom Amiga" is in other words "a
machine that beats everything out there like the Amiga originally did".

I don't think that it's possible to get something that is so much better right
now. The question would be "what shall the new motherboard do better than
what's allready on the market"?

I really don't got a clue.

There are a few things missing though wich I want to see happen. The fantasy-case
is one of them. Cool cases and keyboards wich are Amiga-specific. Right now I'm
using a standard PC-tower and a ugly Fujitsu Siemens keyboard (allthough quite
nice to type on). I wouldn't mind having a new A4000-style keyboard and I would
"love" to have a tower looking as cool as the A4000T did. I really thought that
it had a nice design and I still think so. Besides, didn't Eytech show some
examples of a possible design for a Tower for the A1?

Another thing would be what was descussed on your ML many years ago about runing
old games transparently as if it was an old Amiga. I still think that this is
a nice idea. It can be made by making a script for E-UAE, adding an icon and make
it launch a disk or even a harddrive with a WHDLoad-installed game. What needs
to be fixed regarding that is the way one makes it work. It's a bit geeky to
say the least. Even though I would prefer to be able to install a game on the
harddrive and run it from WB (without noticing E-UAE) a GUI for E-UAE will
probably do.

I also hope that Eyetech will get something new out by the time AOS 4 final is
released. Either a motherboard based on Atricia P or more A1 XE motherboards.

Is this realistic?

@Atheist

Quote:

Did you read post 14 and 17?

No, must have escaped my attention.

Why would you want to "re-invent" chip ram? It is the worst possible feature. Any modern graphics chip can use normal main memory for a lot of things (either through AGP, or through remapping features like the PCI GART in the Radeons) via busmastering, and has local memory for specific, high speed tasks.

Screen dragging would require special precausions in the graphics chip and has nothing to do with an extra CPU for blitting. However, nowadays you would redesign the graphics system to use 3D acceleration and could use screen dragging as well as screen rotating and what have you. An extra CPU for that would require access to the graphics card local video memory.

What you essentially want to see is double the price for a machine to be able to introduce screen dragging?

_________________
Seriously, if you want to contact me do not bother sending me a PM here. Write me a mail

@hotrod

Quote:

I don't think that it's possible to get something that is so much better right now. The question would be "what shall the new motherboard do better than what's allready on the market"?

The plain and simple truth is that nothing that we can do cannot be done better by 'them'. Period. This will remain this way for the forseeable future. This is not any lack of "dreams" or imagination, but humility and knowing one's personal frontiers.

_________________
Seriously, if you want to contact me do not bother sending me a PM here. Write me a mail

@Atheist
>Is it really that expensive to get a company to make these, that could be interfaced with todays
>GPUs? Unfortunately, we've already had a MB designed now. But maybe, when Eyetech is
>rolling in dough, they could give it a shot?


Uhm, nearly everything that Agnus and/or other "custom chips" did already is present in today's PCs. DMA controllers andsystem clocks sounds like Northbridge stuff. Blitter is in the graphics chips from ATI, Nvidia, Matrox, etc.

Chip RAM can be compared to the memory on the graphics card, or memory on sound, ethernet, etc. cards. Seperating these chunks of "chip RAM" to seperate cards to be used exclusively by those cards is even better than a unified "chip RAM" concept, as each of the cards' chips can be doing stuff in parallel each using its own exclusive memory instead of each chip waiting its turn to talk to unified chip RAM, where only one can access memory at a time.

There's really no need for "custom" chips or chip RAM today. I'd like to see some features of the Northbridge we can use today updated, so that we can barely keep up with PCs. We're missing PCI Express, faster memory support, etc. and I wouldn't be unhappy if we saw some more integration of peripherals like gigabit ethernet, firewire, sound, etc. into a single chip with the Northbridge stuff already in Articia. They may do this at some point, and when they do it won't be a "custom chip", it'll be a new off-the-shelf standard product they want to sell to as many companies as possible. It would be coprorate suicide to make a chip and then only sell it to us so it could be considered "custom".

Chips are really expensive. Just the mask set will cost hundreds of thousands of dollars, and those are just the equivalent of camera film negatives, used to expose the wafer during manufacturing. You also have to pay engineers to design, buy software for the engineers to use, buy kits of pieces the engineers assemble together to form completed circuitry, pay test engineers to develop methods to verify the thing works, and then pay test engineers to perform that testing.

_________________
All glory to the Hypnotoad!