Single Board Computer X86: History Of Computing

Knowing the single board computer x86 is knowing the basis of the processors in the still brief history of computing. Millions of processors sold throughout history implement it, and it has also been a design that gave rise to the success of companies such as Microsoft and Intel, as well as the concept of the personal computer. Single board computer x86 is a technological milestone worth analyzing and studying.

Single board computer x86 is one of the most important creations in the world of hardware, and it is essential to understand the history of computing and its current state. 

A little history: Intel 4004

In 1971, Intel presented the first commercial processor on the market, the 4004. It was a model with a 4-bit bus designed for use in calculators, ancient but considered the first stone of everything that would be presented in that decade. It was followed by the Intel 8008 and 8080 in 1972 and 1974, respectively, with an instruction set designed by Datapoint Corporation and intended for use in the most advanced calculators.

After these models, Intel began a huge project with which it sought to reinvent the world of processors. In 1975, it began the designs of the 32-bit iAPX 432 (Intel Advanced Processor Architecture), with improvements in multitasking and memory management compared to the 8000 family, being an architecture designed for object-oriented programming and able to manage multiple processes simultaneously. 

Intel 8086, 1979

In May 1976, they began to develop a new processor that began to be sold a couple of years later, in mid-1978. The Intel 8086 included backward compatibility with the software of the 8008 and 8080, using 3-micron transistors and a frequency of up to 4.77 MHz. Most important was introducing a new set of instructions designed by Intel and called x86-16. It was a small success for Intel, and although it was powerful for the time, its price was considered excessive. 

A few months later, on July 1, 1979, the Intel 8088 was presented, a cheaper version of the 8086 with practically all its features, and which had its main ally in IBM. The IBM PC 5150, considered the first mass-market personal computer (‘PC’) in history, used the 8088 to the detriment of other competing processors due to its high availability, ease of programming, and low price. The IBM PC, which cost around $3,000, was a bestseller since its launch on August 12, 1981.

Then, Intel positioned a new family of chips as its main product: the ‘ single board computer x86‘ processors, defined as using the same set of instructions that, although it has evolved over the years, continues to use many of the original features.

Intel 80286, 1982

The 8086 is considered the first single-board computer x 86 in histories, although the model that catapulted Intel’s fame was the 8088. It was followed by the 80186, 80286, 80386, and 80486, better known as 186, 286, 386, and 486, which were presented throughout the 80s and early 90s within equipment encompassed under the concept that was created in the IBM PC, which was a complete success. 

Since the first single-board computer x86 in the seventies, Intel has been in charge of moving the baton of the main CISC architecture in history. However, it is not the only one: Motorola, DEC, and especially IBM have presented CISC processors throughout history. 

Intel single board computer x86 and its main features

The company’s long experience has resulted in multiple and notable changes in many aspects, ranging from the improvement in the size of the data buses (16, 32, and now 64 bits) to multiple new instructions added that have been adapted to the new uses of technology

For example, the 8086 did not have floating point operations, so Intel created a mathematical coprocessor that performed these operations: it was a small extension on x86 called x87 and launched on the market under the names 8087, 80187, 80287, 80387, 80487 and the last 80587, already in the mid-90s. 
For a couple of decades, all commercial single-board computers x86 processors have included instructions to operate in floating point. Hence, an additional coprocessor is only necessary for specific uses in which great computing power is required, such as with systems like NVidia Tesla.