SuperCPU is an almost entirely self-contained computer, which runs at 20 MHz, leaving the C64 or C128 to run internally at its normal 1 MHz or 2 MHz mode. SuperCPU uses the C64/128 as essentially an I/O adaptor and video RAM,runs at 20 MHz unless software instructs it to communicate with video RAM or I/O devices, or tells it to manually slow down to match the C64 or C128.
SuperCPU employs a one-byte write cache, allowing it to disconnect from C64 or C128 entirely. This allows the programmer to perform a write operation and then immediately go do something else, without waiting, while the cache handles the actual write operation to I/O or video RAM. This applies so long as the programmer spaces his/her writes out so that they occur at least 20 cycles apart, SuperCPU always slows down when reading from an I/O device, and does not use the cache.
The write cache allows SuperCPU code to treat the C64 or C128 as 1 MHz I/O and video memory bus. This allows more video and sound effects than the C64 or C128 would normally be capable of (you can normally write to memory or I/O at about 50 KB/sec).
WDC 65C816 is 65K family CPU which runs emulating 65C02 or 16-bit native mode. Emulated 65C02 replaced 6502 illegal opcodes by extended command set, in exchange for those illegal opcodes, most of 6502 instructions have been upgraded to allow them to take advantage of the full 16 MB of available memory, if present.
SuperCPU comes with 256 KB of 20 ns "cache" RAM, which serves as base memory. User-installed extended memory beyond this is 70 ns or faster.
In 20 MHz mode, the SuperCPU corrects for the 70 ns speed by imposing occasional memory wait states when dealing with extended memory. On average, these wait states incur up to a 5% speed penalty. It is possible to avoid wait-states with carefully written code. With Turbo mode turned off, code runs at the same speed as on stock machines, wherever it is located in the 16 MB address space, and perform block-move commands at a rate of seven clock cycles per byte, regardless of which part of memory that's being dealt with. With Turbo mode turned on, block-moves to the first 256 KB are executed at 7 cycles per byte, and code in this region runs at about 2.71 Million to 10 Million instructions per second (MIPS). Internal memory hacks that take the C64 or C128 beyond 64 or 128 KB are not recognized by the Super CPU. Final note should be that SuperCPU128 design forgot Z80 implementation in Commodore 128 and when is turned on, Z80 inside C128 is not accessible. SuperCPU64 used with Commodore 128 forces computer start directly into Commodore 64 mode.