32, 64, 86: Chip Numbers Explained

MicTig asked the Answer Line forum why we call 32-bit PCs x86 instead of the more descriptive x32. After all, we call their 64-bit equivalents x64.

The 32-bit and the 64-bit chips that run Windows-capable PCs use what's called the x86 instruction set. But since the x86 label was in common use before 64-bit systems hit the market, the later variation on the standard became known as x64, although the correct term is x86-64.

But why is it x86 at all? The explanation goes back to the pre-Pentium days when processors had numbers rather than market-driven brand names. Here's a quick history of the chips that eighty-sixed the standard:

8086: This was Intel's first 16-bit chip, and the beginning of the line. Released in 1978, It was designed to replace Intel's 8-bit 8080, but it wasn't backward compatible as we would understand that term today. You couldn't run 8080 code on an 8086.

8088: Introduced in 1979, this was basically an 8086 with an 8-bit bus (16-bit on the inside, 8-bit on the outside). The result was a slower but cheaper chip for slower but cheaper computers. The original IBM PC and PC-XT were 8088-based.

80286: This 16-bit chip, introduced in 1982, added multitasking and retained 8086 backward compatibility. Unfortunately, it couldn't multitask while maintaining that compatibility. You could only multitask while running an OS and applications written for the 286. These never materialized--or at least not in time. But it was faster than the 8086, and was the basis for IBM's popular PC-AT.

80386: This one changed everything. The first 32-bit chip in the X86 series, it was also the first that multitasked without sacrificing backward compatibility. But it took nearly a decade after its 1985 release--by which time it was no longer around--for a popular OS (Windows 95) to really use its capabilities.

80486: Released in 1989, this was basically a 386 with an integrated floating point processor and an onboard cache.

Pentium: Intel discovered the hard way (by losing a court case) that they couldn't trademark a number, so they called the would-be 80586 the Pentium. Aside from the spiffy new name, the 1993 processor added superscalar architecture, which allows the chip to process more than one command at a time.

There have been plenty of other processors since then, from both Intel and AMD. But they've pretty much stuck to names instead of numbers.

Read the original thread at http://forums.pcworld.com/message/178140.

Add your comments to this article below. If you have other tech questions, email them to me at answer@pcworld.com, or post them to a community of helpful folks on the PCW Answer Line forum.

Image: superstar (Flickr/CC)

To comment on this article and other PCWorld content, visit our Facebook page or our Twitter feed.
Shop Tech Products at Amazon