Why did Atari produce TWO 8 bit systems?

[downix]

Had Atari progressed as Jay Miner and company had envisioned, the Atari 400/800 would have debuted as game machines in 1979, followed by the Amiga circa 1983/84.

 

Don't confuse the Atari Amiga (also called Mickey) with the final result, the Commodore Amiga. While the concepts and core chip designs are the same, Commodore implimented many changes to streamline integration with their technology plants. (The CIA chips, for example) Frankly, I feel that both sides lost, and gained, from the end deal. Atari's Amiga would have lacked much of the OS that made Commodores Amiga prized even today. (Workbench was started at Commodore in '83 for the original C65 68000-based C64 replacement, not to be confused with the early 90's C65 project which used the same custom chips, but not the advanced 68000 CPU. Commodore also licensed TripOS for use as the DOS)

 

Would it have been better, or worse? Who knows. I do know that there is something special about the custom chips, and that I hope to see someone take that vision to the next level.

[slapdash]

If all 3 are 8 bit systems, then that doesn't make sense to me. How can a 5200 be superior to a 2600? Is this based solely on factors such as RAM, Processing speed and other factors?

 

Even though I think this question has been adequately answered, I'll offer this analogy -- are all V8s the same?

[kisrael]

Alot of  classic 8-bit systems used the 6502 CPU, even Commodore 1541 and 1571 disk drives had one! funny but true

Yeah, I heard there were some heavy-calculation programs that let you use the disk drive's processor...in some ways it was almost as 'smart' as the main unit...like w/ videogames, it's the extra chips to do tasks (as well as clock speed) that really makes the difference, much more than 8-bits.

 

Basically, with an 8 bit system, it's easy to go up to integer values of up to 255 (sound familiar? That shows up in a lot of NES games...) and just somewhat trickier slower to deal with larger numbers. With a 16-bit system, it's easy to go up to up to 65,535 (which may also sound familiar to some). Looks like a 32 bit system can trivially go up to over 4 billion, but in practice you might use fewer bits for most things. Because those number limits are "natural", it also ties in with how big a memory you can easily work with.

 

But there's rather little an 8 bit system CAN'T do that 16 bit system can; it might take just a bit longer. The special purpose chips for graphics and what not are much more important.

[Tsukasa]

You are. There is no 68000 in the Lynx. It uses a custom 6502 for it's CPU.

 

Oh, yes (what was I thinking )

 

But I'm very sure it's a 16bits 6502, in other words a 65816 (like in the SNES)

 

The SEGA Genesis contains a 68000

 

The genesis also had a Z80 in it.

[Shawn Jefferson]

But I'm very sure it's a 16bits 6502, in other words a 65816 (like in the SNES)

 

The Lynx is a plain old 65C02, an 8-bit processor, running at 4Mhz. There are a few more instructions than what you are used to on the 6502.

 

I always wondered why the Suzy graphics chip was called a 16-bit chip. When speaking about bit width of CPUs, usually that refers to the data bus right? Did the Suzy actually have a 16-bit data bus, or is there some other reason it is called a 16-bit chip?

 

Also, I seem to recall that the 1050 drive has a 65xx series chip in it as well, is that true?

[Gregory DG]

The Lynx is a plain old 65C02, an 8-bit processor, running at 4Mhz.

Well, the "C" is supposed to stand for "Custom" so I don't think it was the "plain old" 65C02.

[NovaXpress]

You can blow a modern gamer's mind by telling them that the XBox is only a 32-bit system. This leads to a discussion of what that actually means. Because of all the marketing hype in the 16-bit era, most gamers believe that more bits=better games and there's nothing more to it than that.

[Tsukasa]

The Lynx is a plain old 65C02, an 8-bit processor, running at 4Mhz.

Well, the "C" is supposed to stand for "Custom" so I don't think it was the "plain old" 65C02.

 

Actually, the "C" stands for CMOS. The 65C02 is the CMOS version of the 6502, it lowers the power consumption and adds a few new instructions but isn't "custom" in any way.

[Gregory DG]

Actually, the "C" stands for CMOS. The 65C02 is the CMOS version of the 6502, it lowers the power consumption and adds a few new instructions but isn't "custom" in any way.

Hmm. That's interesting. In 15 years of reading info about the Lynx, I've never heard that one!

[Tsukasa]

Actually, the "C" stands for CMOS. The 65C02 is the CMOS version of the 6502, it lowers the power consumption and adds a few new instructions but isn't "custom" in any way.

Hmm. That's interesting. In 15 years of reading info about the Lynx, I've never heard that one!

 

Read this page.

[TuzenTCA]

You can blow a modern gamer's mind by telling them that the XBox is only a 32-bit system. This leads to a discussion of what that actually means.

 

What? The XBOX is 32? But the 64 is so primitive and yet it is a 64! I thought that the Dream Cast, the Sony Playstation 2, and the XBOX all had 128 bit graphics processors in them. Am I wrong? I thought it went in translation from 32 to 64 and then to 128. I thought the PS1 was a 64 bit, yet Ben Heckendorn labels it as: "Original Sony 32-bit Sony Playstation 1". Any help on the matter would be appreciated.

 

Thanks,

 

James McAninch

[Adrian M]

 

The genesis also had a Z80 in it.

 

Indeed it did. It was included to handle the sound processing duties as well as allowing SMS games to be played by way of the Power Base Convertor.

[Vigo]

Basically, with an 8 bit system, it's easy to go up to integer values of up to 255 (sound familiar? That shows up in a lot of NES games...) and just somewhat trickier slower to deal with larger numbers.  With a 16-bit system, it's easy to go up to up to 65,535 (which may also sound familiar to some).

 

The 68000, although called a 16 bit chip, is a full 32bit design internally. All its registers are 32 bit wide, and you can easily perform 32bit operations. It just takes, due to the 16 bit data bus, two memory cycles to write a 32bit value into memory.

 

Looks like a 32 bit system can trivially go up to over 4 billion, but in practice you might use fewer bits for most things.   Because those number limits are "natural", it also ties in with how big a memory you can easily work with.

 

You also need many bits to calculate floating point numbers precisely.

 

But there's rather little an 8 bit system CAN'T do that 16 bit system can; it might take just a bit longer.  The special purpose chips for graphics and what not are much more important.

 

Yeah, the SNES is the best example for this. It has a slow 8 bit CPU (65816, yes, it IS an 8 bit cpu!), but most games look much better, than the average Genesis game, despite the fact the Genesis has a much more powerful CPU.

[kisrael]

Looks like a 32 bit system can trivially go up to over 4 billion, but in practice you might use fewer bits for most things.   Because those number limits are "natural", it also ties in with how big a memory you can easily work with.

 

You also need many bits to calculate floating point numbers precisely.

 

Yeah...16-bit era this didn't come up quite so often I suspect?

 

Though I did 16 bit math (fixed fractional, essentially) to get some smooth movement effects in JoustPong...

[EricBall]

You also need many bits to calculate floating point numbers precisely.

Well, yes & no. Floating point numbers are just a different way of expressing a value. If n and m are a number of bits, then the range of different formats is.

 

Unsigned 0 to (2^n)-1

twos complement -2^(n-1) to (2^(n-1))-1

signed fixed point -2^(n-1) to 2^(n-1))-1 + ((2^m)-1)/(2^m)

unsigned floating point 0 to 2^((2^n)-1) * ((2^m)-1)/(2^m)

 

Sorry if I've lost anyone there. Okay, the unsigned is easy to understand. 8 bits has the range 0 to 255. 8 bit twos complement is -128 to 127. Fixed point has two components, a integer portion and a fractional portion. So 16 bits (8 integer, 8 fractional) can represent the signed values -128 to 127+255/256.

 

Floating point also has two components, a fractional portion and a mantissa (or exponent). So an 16 bit unsigned floating point number (8 fraction, 8 mantissa) has the range 0 to 2^255 * 255/256.

 

As you can see, there is nothing implicit with floating point representation which requires additional bits. However, more mantissa bits increase the range of values which can be represented and more fractional bits increase the precision of the value. For example 1/3 would be have to be approximated as 85/256 with an 8 bit fraction. Note: no matter how many bits are used, all numbers can not be represented exactly. There will always be an error.

 

Floating point is very useful when performing multiplication since the range of values is much, much larger than integer or fixed point with the same number of bits. However, floating point addition is much more difficult. Multiplication in general requires significantly more CPU processing to accomplish, so was typically avoided by time-sensitive applications (e.g. games) until the transistor budget was large enough to create dedicated high-speed hardware multiply, divide & floating point sub-processors.

[cwilkson]

The Lynx is a plain old 65C02' date=' an 8-bit processor' date=' running at 4Mhz.[/quote'']

Well, the "C" is supposed to stand for "Custom" so I don't think it was the "plain old" 65C02.

 

Actually, the "C" stands for CMOS. The 65C02 is the CMOS version of the 6502, it lowers the power consumption and adds a few new instructions but isn't "custom" in any way.

 

This is a common source of confusion. It took me like 5 years to finally get it right.

 

I don't know what the Lynx has in it. But the 7800 (and I think the 8-bit computers, including the 5200) has a 6502C. That "C" is for "Custom". The 65C02 as Tsukasa points out is just a CMOS version of the original NMOS 6502. The two chips are functionally different. The 6502C adds a /HALT line to stop the processor in its tracks. Also, the pinout is different. The R/W line is on a different pin. That pin is unused on the 6502 (and 65C02).

 

In short, the 65C02 is a direct low-power pin-for-pin replacement for the 6502. The 6502C is a minor upgrade for the 6502 with a different pinout. It was developed by Atari, or developed specifically for them by someone else.

 

-Chris