CompuServe Thread

From what I've heard of the 65C816 machine, its the long rumored Apple IIx "16-bit" Apple II series machine. The 65C816 processor is very much like the 6502 (in fact, there is a pin-compatible version called the 65C802). It has a 6502 emulation mode, and a pseudo-16 bit mode, similar to the 8088, i.e. internal 16 bit, external 8 bit data path, page registers and the like (not to imply any 8088 compatibility). From a programmer's point of view, you have an extended 6502 instruction set (similar to the 65C02 chips) in which each instruction is overloaded for 8 and 16 bit operation. A bit in the status register controlls whether the instruction performs an 8 or 16 bit operation. No multiply or other advanced instructions are available, though there is one designed to support a yet-to-be-announced coprocessor. The chip supports 24 different addressing modes, and 24 bits of address (the high-order 8 bits are multiplexed with the data bus). All programs must run in a 64K bank, but there are jump and subroutine instructions that allow movement in the full 16M space. Also, one addressing mode permits a form of data addressing that goes beyond the 64K limit of a block. All-in-all, its not an entirely bad replacement for a 6502, and there are some things about it superior to the 8088 (and some inferior). It doesn't compare well to the 68000 at all (only 1 general purpose and 2 index registers, each 16 bits long, no frame pointer register, etc., no protected operation modes, etc.). The rumors about the new Appe machine also tend to claim something in a 7xx by 3xx display, 7 or 8 bits deep. I could not imagine the '816 running this at any reasonable speed, even though an 8 bit display organized as 1 byte/pixel could speed up bit mapping by removing the need for pixel addressing. Also, the rumors put the basic system price in the $1600 range. Today, I laugh loudly.

From what I've heard of the 65C816 machine, its the long rumored Apple IIx "16-bit" Apple II series machine. The 65C816 processor is very much like the 6502 (in fact, there is a pin-compatible version called the 65C802). It has a 6502 emulation mode, and a pseudo-16 bit mode, similar to the 8088, i.e. internal 16 bit, external 8 bit data path, page registers and the like (not to imply any 8088 compatibility). From a programmer's point of view, you have an extended 6502 instruction set (similar to the 65C02 chips) in which each instruction is overloaded for 8 and 16 bit operation. A bit in the status register controlls whether the instruction performs an 8 or 16 bit operation. No multiply or other advanced instructions are available, though there is one designed to support a yet-to-be-announced coprocessor. The chip supports 24 different addressing modes, and 24 bits of address (the high-order 8 bits are multiplexed with the data bus). All programs must run in a 64K bank, but there are jump and subroutine instructions that allow movement in the full 16M space. Also, one addressing mode permits a form of data addressing that goes beyond the 64K limit of a block. All-in-all, its not an entirely bad replacement for a 6502, and there are some things about it superior to the 8088 (and some inferior). It doesn't compare well to the 68000 at all (only 1 general purpose and 2 index registers, each 16 bits long, no frame pointer register, etc., no protected operation modes, etc.). The rumors about the new Appe machine also tend to claim something in a 7xx by 3xx display, 7 or 8 bits deep. I could not imagine the '816 running this at any reasonable speed, even though an 8 bit display organized as 1 byte/pixel could speed up bit mapping by removing the need for pixel addressing. Also, the rumors put the basic system price in the $1600 range. Today, I laugh loudly.