Kinds of Microcomputer

Microcomputers are broadly divided into microprocessors (MPUs or CPUs) that are designed to be connected to external memory or other peripheral devices, and microcontrollers (MCUs) that have on-chip memory and peripheral functions for embedded applications.

The key features in MPUs are expandability and performance, while those in MCUs depend on the application, which means that specifications vary according to the target system.

Since each MCU has a system implemented on a single chip, MCUs are also called single-chip or one-chip microcontrollers.

First, these terms are read "sisk" and "risk", and they are acronyms that stand for the following:
CISC: Complicated Instruction Set Computer
RISC: Reduced Instruction Set Computer

As microprocessors have evolved, processing speeds have increased and instructions have become more complex.
Designers realized that if, for example, there are three instructions representing basic functions A, B, and C, it would be convenient to also have single instructions that can execute complex combined processing such as A + B, B + C, or A + B + C, as well as a lot of addressing. Addressing has also been diversified to include not only direct addressing, where the access address is described as the operand, but also indirect addressing, where the contents of various registers are used as the access address, or the contents of multiple registers are combined and made the access address, etc.
This trend led to the development of the CISC concept. However, although not all instructions are actually frequently used, the number of instructions became very large.
The more instructions there are, the heavier the load on the instruction decoder, whose job it is to analyze the instructions (determine what needs to be done) before they are executed. It is time-consuming to do this analysis by software, and it increases the scale of circuitry to do it by hardware. Moreover, a large circuit scale means that more power is required to deliver the required processing speed. Complex addressing also leads to a heavy processing load during instruction execution.
The RISC concept was introduced to overcome these problems by limiting the instruction set to the basic instructions used most frequently.
Note that the spelling differs from "risk", and thus the meaning is not "risky microcomputer."

Generally, it indicates the number of bits (data width) that the device can process at one time.

For example, while an 8-bit microcontroller perform operations on only 8 bits at a time, a 32-bit microcontroller can do this on up to 32 bits at time.
Although a 32-bit microcontroller can process data faster, this performance capacity is wasted unless the target system also supports a similarly large number of bits per operation. In other words, a large bit count in the microcontroller does not necessarily mean that a system using that device will run faster.

Some microcontroller versions have smaller data widths for input and output than for internal processing, such as 32-bit microcontrollers that have a 16-bit external bus width.

Different MCU models feature on-chip PROM, flash memory, or mask ROM, and some are ROMless.

PROM and flash memory can be programmed by users via a programming tool (programmer), and they are often used for testing, evaluation, and small-scale production.

PROM versions include once-writeable OTP (One-time Programming) and rewriteable EPROM types.
Currently, the most common type of programmable on-chip memory is flash memory.

Mask ROM versions are semi-custom products that are manufactured via a process that uses a dispersed mask (for LSI manufacturing), and user programs are written to this ROM before the product is shipped.

ROMless versions connect to external ROMs that are programmed for various applications. These MCUs are designed for systems that can switch among various external and replaceable ROMs, such as for testing, evaluation, etc.

A flash microcontroller is a microcontroller that includes flash memory, which is used to store programs.

The biggest difference is that flash microcontrollers can be re-programmed after they are shipped.
Data written to flash memory can be electrically erased, allowing new data to be written.
For older types of flash memory, which use two power supplies (in addition to the power supply for VDD, a separate power supply is provided for VPP for programming), the program can be overwritten by using a dedicated flash programmer (FlashPro, etc.).
More recent types of flash memory are single power supply versions (which do not need a separate power supply for programming). Besides dedicated flash programmers, there are also other ways to program/reprogram and write new data to these devices.
Also, among the microcontrollers with single power supply type flash memory that can be programmed after shipment, there are some in which part of the flash memory used to store data instead of an EEPROM.
Thus, flash microcontrollers can serve a wide range of applications.
more product families that use only flash memory and no longer include mask ROM.

Basically, that concept is correct. Microcontrollers derived from conventional products that include mask ROM products and PROM products are not categorized as All Flash.
All Flash microcontrollers achieve a power consumption equivalent or lower than that of mask ROM products.

Among All Flash microcontrollers, low-pin-count microcontrollers are products such as the 78K0S (78K0S/Kx1+) Microcontrollers that have approximately 30 or fewer pins.