X86 Assembly/X86 Family

From Wikibooks, open books for an open world
< X86 Assembly
Jump to: navigation, search

The term "x86" can refer both to an instruction set architecture and to microprocessors which implement it. The name x86 is derived from the fact that many of Intel's early processors had names ending in "86".

The x86 instruction set architecture originated at Intel and has evolved over time by the addition of new instructions as well as the expansion to 64-bits. As of 2009, x86 primarily refers to IA-32 (Intel Architecture, 32-bit) and/or x86-64, the extension to 64-bit computing.

Versions of the x86 instruction set architecture have been implemented by Intel, AMD and several other vendors, with each vendor having its own family of x86 processors.

Intel x86 Microprocessors[edit]

8086/8087 (1978)
The 8086 was the original x86 microprocessor, with the 8087 as its floating-point coprocessor. The 8086 was Intel's first 16-bit microprocessor with a 20-bit address bus, thus enabling it to address up to 1 Megabytes, although the architecture of the original IBM PC imposed a limit of 640 Kilobytes of RAM, with the remainder reserved for ROM and memory-mapped expansion cards, such as video memory. This limitation is still present in modern CPUs, since they all support the backward-compatible "Real Mode" and boot into it.
8088 (1979)
After the development of the 8086, Intel also created the lower-cost 8088. The 8088 was similar to the 8086, but with an 8-bit data bus instead of a 16-bit bus. The address bus was left untouched.
80186/80187 (1982)
The 186 was the second Intel chip in the family; the 80187 was its floating point coprocessor. Except for the addition of some new instructions, optimization of some old ones, and an increase in the clock speed, this processor was identical to the 8086.
80286/80287 (1982)
The 286 was the third model in the family; the 80287 was its floating point coprocessor. The 286 introduced the “Protected Mode” mode of operation, as opposed to the “Real Mode” that the earlier models used. All subsequent x86 chips can also be made to run in real mode or in protected mode. Switching back from protected mode to real mode was initially not supported, but found to be possible (although relatively slow) by resetting the CPU, then continuing in real mode. Although the processor featured an address bus with 24 lines (24 bits, thus enabling to address up to 16 Mebibytes), these could only be used in protected mode. In real mode, the processor was still limited to the 20-bits address bus.
80386 (1985)
The 386 was the fourth model in the family. It was the first Intel microprocessor with a 32-bit word. The 386DX model was the original 386 chip, and the 386SX model was an economy model that used the same instruction set, but which only had a 16-bit data bus. Both featured a 32-bits address bus, thus getting rid of the segmented addressing methods used in the previous models and enabling a "flat" memory model, where one register can hold an entire address, instead of relying on two 16-bit registers to create a 20-bit/24-bit address. The flat memory layout was only supported in protected mode. Also, contrary to the 286, it featured an "unreal mode" in which protected-mode software could switch to perform real-mode operations (although this backward compatibility was not complete, as the physical memory was still protected). The 386EX model is still used today in embedded systems,
80486 (1989)
The 486 was the fifth model in the family. It had an integrated floating point unit for the first time in x86 history. Early model 80486 DX chips were found to have defective FPUs. They were physically modified to disconnect the FPU portion of the chip and sold as the 486SX (486-SX15, 486-SX20, and 486-SX25). A 487 "math coprocessor" was available to 486SX users and was essentially a 486DX with a working FPU and an extra pin added. The arrival of the 486DX-50 processor saw the widespread introduction of fanless heat-sinks being used to keep the processors from overheating.
Pentium (1993)
Intel called it the “Pentium” because they couldn't trademark the code number “80586”. The original Pentium was a faster chip than the 486 with a few other enhancements; later models also integrated the MMX instruction set.
Pentium Pro (1995)
The Pentium Pro was the sixth-generation architecture microprocessor, originally intended to replace the original Pentium in a full range of applications, but later reduced to a more narrow role as a server and high-end desktop chip.
Pentium II (1997)
The Pentium II was based on a modified version of the P6 core first used for the Pentium Pro, but with improved 16-bit performance and the addition of the MMX SIMD instruction set, which had already been introduced on the Pentium MMX.
Pentium III (1999)
Initial versions of the Pentium III were very similar to the earlier Pentium II, the most notable difference being the addition of SSE instructions.
Pentium 4 (2000)
The Pentium 4 had a new 7th generation "NetBurst" architecture. Pentium 4 chips also introduced the notions “Hyper-Threading”, and “Multi-Core” chips.
Core (2006)
The architecture of the Core processors was actually an even more advanced version of the 6th generation architecture dating back to the 1995 Pentium Pro. The limitations of the NetBurst architecture, especially in mobile applications, were too great to justify creation of more NetBurst processors. The Core processors were designed to operate more efficiently with a lower clock speed. All Core branded processors had two processing cores; the Core Solos had one core disabled, while the Core Duos used both processors.
Core 2 (2006)
An upgraded, 64-bit version of the Core architecture. All desktop versions are multi-core.
i Series (2008)
The successor to Core 2 processors, with the i7 line featuring Hyper-Threading.
Celeron (first model 1998)
The Celeron chip is actually a large number of different chip designs, depending on price. Celeron chips are the economy line of chips, and are frequently cheaper than the Pentium chips—even if the Celeron model in question is based off a Pentium architecture.
Xeon (first model 1998)
The Xeon processors are modern Intel processors made for servers, which have a much larger cache (measured in megabytes in comparison to other chips' kilobyte-sized cache) than the Pentium microprocessors.

AMD x86 Compatible Microprocessors[edit]

Athlon is the brand name applied to a series of different x86 processors designed and manufactured by AMD. The original Athlon, or Athlon Classic, was the first seventh-generation x86 processor and, in a first, retained the initial performance lead it had over Intel's competing processors for a significant period of time.
Turion 64 is the brand name AMD applies to its 64-bit low-power (mobile) processors. Turion 64 processors (but not Turion 64 X2 processors) are compatible with AMD's Socket 754 and are equipped with 512 or 1024 KiB of L2 cache, a 64-bit single channel on-die memory controller, and an 800 MHz HyperTransport bus.
The AMD Duron was an x86-compatible computer processor manufactured by AMD. It was released as a low-cost alternative to AMD's own Athlon processor and the Pentium III and Celeron processor lines from rival Intel.
Sempron is, as of 2006, AMD's entry-level desktop CPU, replacing the Duron processor and competing against Intel's Celeron D processor.
The AMD Opteron is the first eighth-generation x86 processor (K8 core), and the first of AMD's AMD64 (x86-64) processors. It is intended to compete in the server market, particularly in the same segment as the Intel Xeon processor.