Intel 8086

Intel 8086 - A piece of microprocessor of sixteen bits used in GCV early of IBM. Intel 8088 was a version with an external data bus of eight bits.

Intel 8086 was based on the design of Intel 8080 and Intel 8085 (it was source compatible with the 8080) with a similar whole of register, but was increased with 16 bits. The unit of interface of bus supplied the jet of instruction to the module execution by a queue of prefetch of 6 bytes, thus the effort and the execution was convergent - a primitive form of drain (8086 varied instructions from 1 to 4 bytes).

It comprised four general registers of 16 bits, which could also be reached as eight registers 8 bits, and four index registers of 16 bits (indicator of pile including). The registers of data were employed often implicitly by instructions, complicating the attribution of register for provisional values. It comprised the ports with 8 bits bit of the input-output 64K (or the 32K 16) and fixed directed interruptions. There were also quadripartite registers which could be placed starting from the index registers.

The registers of segment allowed the central processing unit treatment to reach 1 megohm from memory in an odd way. Rather than just of the bytes absent suppliers, like the majority of the segmented processors, the 8086 really shifted the registers of segment left 4 bits and added them to the address. Consequently, the segments covered, and were possible for him to have two indicators with the same point of value at two different places of memory, or two indicators with various values indicating the same place. The majority of the people consider this a design damaged by brain.

Although it was mainly acceptable for assembly language, where the order of the segments was complete (it could even be useful then), in languages moreover high level it because constant confusion (for example close of far from the indicators). Still worse, this returned increasing the space of addressing to more than 1 difficult megohm. A posterior version, Intel 80386, increased the design with 32 bits, and fixed segmentation, but required additional modes (removing the new devices) for compatibility, and maintains architecture awkward. In fact, with the right assembler, the code written for the 8008 can still be run on Intel most recent 486.

Intel 80386 new additional codes operation of a kludgy fashion similar to Zilog Z80 and Zilog Z280. Intel 486 full additional drains, and doubly of clock (like Zilog Z280).

Thus why did it make IBM it chose the 8086 series when the majority of the alternatives were so better? Apparently IBM 's have engineers wanted to employ Motorola 68000, and it was employed later in the forgotten computer of laboratory of instruments 9000 of IBM, but IBM already had rights to manufacture the 8086, in exchange to give at Intel the rights to its designs of bubble memory. Apparently IBM employed 8086s in the unit of word processing of IBM Displaywriter.

Other factors were the version with 8 bits of Intel 8088, which could employ Intel existing 8085 typewriting of the components, and made it possible the computer to be based on a modified design 8085. 68000 components were not largely - available, although it could employ Motorola 6800 components until a degree.

The bubble memory of Intel were on the market during one moment, but faded far while the best and cheaper technologies from memory arrived.