CU Amiga -- History 680x0 Family When the Amiga was first launched, the Motorola family of microprocessors was the obvious choice. With the 68000, Motorola had started from scratch to create a true 16 bit processor. Unlike Intel, makers of the 80286, 386, 486 and Pentium family, Motorola decided to abandon any plans for "backwards compatibility". There was no need for convoluted tricks such as Intel-style memory addressing: an Intel processor can have several addresses for the same memory locations. There would be no hardware imposed limits of 640K. The 68000 microprocessor would be good at supporting multitasking and as it saw memory as one flat area, that alone made programming a lot easier. At launch therefore, the 68000 was an impressive chip and the ideal choice to build a new computer around. As the designers of the Amiga new, better Motorola chips would follow in the form of the 68020 and 68030. Improved performance, 32 bit power, higher operating speeds, support for Floating Point Units and Memory Management.. things were looking good. The 68040 was even more stunning, with an integral Floating Point Processor and impressive speed. The 68050 chip never appeared, but the 68060 doubled the speed over the 68040 and crammed in some new features. Sadly, the limitations of the 680x0 family were now obvious. The decision to make the 680x0 processors "do everything" devices rather than lean, mean RISC machines meant that improving performance further was getting extremely difficult. There are only so many components which can be fabricated onto a silicon wafer in a reliable and cost-effective way. The 68060 was pushing this limit and was therefore destined to be the last of the series. In the meantime, the Intel family had continued to be improved. The near total dominance of so-called "WIntel" computers meant there was huge demand for increased performance and huge amounts of money at stake. The Motorola series may have been in great demand by Amiga and Apple users, but the numbers were small and shrinking compared to PC users. Intel chip prices fall on a monthly or weekly basis. whilst Motorola chips are actually in short supply: a few months ago, 68030 processors were unavailable. The Intel Pentium chip is a very powerful piece of silicon, and the entry level chip now runs at 100Mhz (or 120Mhz by the time you read this) much faster than even a top of the range 68060. Custom Chips Brute processing power was never the sole point of the Amiga. The heart of the system, and the reason it out-classed rival 68000 machines and ran rings around the PCs of the day, was the custom hardware. The Amiga was really a multiple processor device. The Motorola chip ticked away in the centre, running programs and holding everything together, but graphics and sound were taken care of independantly. The PAD (Paula, Agnus and Denside) custom chips were stunning when the Amiga first appear. A separate processor, the Copper, looked after the Amiga display, and made possible multiple simultaneous screen modes, smooth scrolling and thousands of colours. No one had seen anything like it. The AGA chipset provided a few tweaks, improved speed and added more colours. Unlike PCs, the Amiga was always geared towards video. Displaying the Amiga Workbench on a colour portable (ideal for home use) and using a Genlock to add DIY subtitles were features even today other machines have real problems with. New and Improved. Initially, the video friendly aspects of the Amiga display were winning combinations. Few people could justify a dedicated computer monitor, and those that did were rewarded only with a slightly crisper image rather than new resolutions. Now things are different. The price and performance of SVGA monitors make them wide spread, and it makes sense to take advantage of the larger screen resolutions possible. These days, a 640 by 256 display is simply too small for anything other than simple games. The minimum should be 800 by 600, preferably 1024 by 768 and the maximum should be very large indeed. Doom, the game which was probably instrumental in almost killing the Amiga as a games machine, exploited the one thing which PC graphics cards were very good at: chunky pixel displays. Whereas the Amiga loves bitplanes and playfields, the PC likes dealing with individual pixels -- this makes drawing textures very quick, and that of course means 3D games like Doom, Duke Nukem, Descent and Quake. Any new graphics system for the Amiga needs to be able to operate in a Chunky way. If the AGA chipset is only going to be tinkered with and sped up, it won't make a big difference. Instead, for a good indication of how things could be done, look at the CyberGraphics system which is used to emulate ordinary Amiga operations on third party graphics cards. CyberGraphics supports practically all Amiga Workbench programs with a host of features. A new 3D library is due soon which could be one of the most important additions to Amiga graphics in recent years. More Power? The end of the Motorola 680x0 family makes it obvious that a new processor family is required ASAP to carry the Amiga forward. Whilst the Amiga was in suspended animation, Apple progressed from the 680x0 to the PowerPC range. This move was a logical choice at the time because the PowerPC is exceptionally good at emulation, 680x0 emulation in particular, which makes porting software -- such as the operating systems -- considerably quicker. Companies such as Phase5 are keen on this solution, and have promised accelerator cards based on the PowerPC "real soon now." On the downside, the PowerPC doesn't appear to be as fast as many people initially hoped, and particularly with the success and relatively low-price of Pentium systems, it no longer appears the panacea it once did. But don't fret: there are plenty of alternatives to the PowerPC to choose from. The DEC Alpha processor has been mentioned, and it would be remiss not to mention the latest ARM processors from once-rivals Acorn: they are extremely fast and can be run using the waste heat generated by a Pentium. VISCorp's role in the development of the new Amiga seems to limited to their set-top box and an improved version of an A1200. They lack the will to build a new, PC-killer machine, and who could blame them? Trying to find a mass market for a machine which offers slightly better performance than a kitted out Pentium PC but at considerable increase in price would be an expensive and fruitless task. Better for VISCorp to stick to what they know, and oversee the licensing of the Amiga technology to companies such as MacroSystems and Phase5. If common sense provails, the new killer Amigas which will appear, will come from different manufactures but all carry an "Certified Amiga Compatible" sticker. Perhaps there will be different levels of compatibility, depending on the requirements of the software. The Amiga itself has reached the most important junction. One path leads to the set-top box: a friendly, mass-market and, compared to PCs, underpowered domestic applicance. If it catches on the way VISCorp want, it will however create an explosion in the amount of software required. The other path, a tortuous and narrow trail, leads to new technology, ludicrous amounts of speed and niche markets in the video and graphics fields. With the undecided nature of future hardware, VISCorp are unable to become a version of "MicroSoft" and simply bundle the AmigaOS with every clone machine made. Instead, they hope to have a supervisory role as keeper of standards. What the Amiga is good at * Good Operating System * Video friendly graphics * Affordable What the Amiga is currently bad at: * Updating the processing power * Supporting hardware standards such as ISA or PCI * Running other operating systems such as UNIX or Win95 * Software market drying up * Sound and graphics now five years behind PC -- Points of View The Amiga as we know it is dead. Commodore smothered it, and with it the hopes of any easy passage to a Next Generation machine. Instead we are left with a nice (although single user) Operating System and a video-orientated chipset which is out of date and cannot hold a candle to the latest PC display cards or PlayStation/Saturn consoles. Yes, of course, a new Amiga with a kick-ass processor and state-of-the-art graphics would be nice: but it's not going to happen. The effort and expensive required to build such a machine -- and building one from scratch is what we are talking about here -- is way too much. Who would buy something like this anyway? It would be two times as expensive as a Pentium based PC and lack the software to take advantage of the hardware. It's a dead end as a mass market machine, and although it would find a niche for tech heads and video fans, it wouldn't sell in sufficient numbers to be interesting. To re-cap, what we have now is a nice OS and some Video chips. Which is exactly what VisCorp want of course, for their consumer orientated set-top box. Yes, Commodore tried it with the CDTV, but this time there is the Internet and CableTV companies to help it along. The right sort of deal would see the box being supplied for free to your home, complete with all sorts of home shopping and pay-per-view systems. The Amiga would pay for itself in months. It just might work -- and if it does, it will spawn brand new desktop systems. Why? Because if there is a Amiga (in some shape or form) in a large percentage of homes, there will be a need for software. That means authoring systems and that means proper Amiga computers again. It also means a potential software market bigger than ever before: imagine if you could sell software to owners of video recorders -- now you are getting an inkling of the size of the market VisCorp are playing for. So, the Amiga is dead -- but it also has the best ever chance of coming back. Better than AT. Better than Commodore even. Trying to sell it as a home computer is doomed to failure, but making it a domestic Internet/Smart Telephone/TV box is the Amiga's last best chance for a peace of the future.