This articles covers the floating-point unit (FPU) contained in the processor, which was formerly a separate external math coprocessor in the 386 and older chips. Older central processing units designed by Intel (and cloned by other companies) used an external math coprocessor chip.

However, when Intel introduced the 486DX, it included a built-in math coprocessor, and every processor built by Intel (and AMD and Cyrix, for that matter) since then includes a math coprocessor. Coprocessors provide hardware for floating-point math, which otherwise would create an excessive drain on the main CPU.

Math chips speed your computer's operation only when you are running software designed to take advantage of the coprocessor. All the subsequent fifth- and sixth-generation Intel and compatible processors (such as those from AMD and Cyrix) have featured an integrated floating-point unit.

Math chips (as coprocessors sometimes are called) can perform high-level mathematical operations—long division, trigonometric functions, roots, and logarithms, for example—at 10–100 times the speed of the corresponding main processor.

The operations performed by the math chip are all operations that make use of noninteger numbers (numbers that contain digits after the decimal point). The need to process numbers in which the decimal is not always the last character leads to the term floating point because the decimal (point) can move (float), depending on the operation.

The integer units in the primary CPU work with integer numbers, so they perform addition, subtraction, and multiplication operations. The primary CPU is designed to handle such computations; these operations are not offloaded to the math chip.

The instruction set of the math chip is different from that of the primary CPU. A program must detect the existence of the coprocessor and then execute instructions written explicitly for that coprocessor; otherwise, the math coprocessor draws power and does nothing else.

Fortunately, most modern programs that can benefit from the use of the coprocessor correctly detect and use the coprocessor. These programs usually are math intensive: spreadsheet programs, database applications, statistical programs, and graphics programs, such as computer-aided design (CAD) software.

Word processing programs do not benefit from a math chip and therefore are not designed to use one. Table below summarizes the coprocessors available for the Intel family of processors.

Processor | Coprocessor |
---|---|

8086 | 8087 |

8088 | 8087 |

286 | 287 |

386SX | 387SX |

386DX | 387DX |

486SX | 487SX, DX2/OverDrive |

487SX | Built-in FPU |

486SX2 | DX2/OverDrive |

486DX | Built-in FPU |

486DX2 | Built-in FPU |

486DX4/5x86 | Built-in FPU |

Intel Pentium/Pentium MMX | Built-in FPU |

Cyrix 6x86/MI/MII | Built-in FPU |

AMD K5/K6/Athlon/Duron | Built-in FPU |

Pentium II/III/Celeron/Xeon | Built-in FPU |

Pentium 4 | Built-in FPU |

Athlon 64 | Built-in FPU |

Itanium/Itanium II | Built-in FPU |

FPU = Floating-point unit |

^{[1]}The 487SX chip is a modified pinout 486DX chip with the math coprocessor enabled. When you plug in a 487SX chip, it disables the 486SX main processor and takes over all processing.

^{[2]}The DX2/OverDrive is equivalent to the SX2 with the addition of a functional FPU.

Although virtually all processors since the 486 series have built-in floating-point units, they vary in performance. Historically, the Intel processor FPUs have dramatically outperformed those from AMD and Cyrix, although AMD and Cyrix are achieving performance parity in their newer offerings.

Within each of the original 8087 group, the maximum speed of the math chips varies. A suffix digit after the main number, as shown in Table below, indicates the maximum speed at which a system can run a math chip.

Part | Speed |
---|---|

8087 | 5MHz |

8087-3 | 5MHz |

8087-2 | 8MHz |

8087-1 | 10MHz |

287 | 6MHz |

287-6 | 6MHz |

287-8 | 8MHz |

287-10 | 10MHz |

The 387 math coprocessors and the 486 or 487 and Pentium processors always indicate their maximum speed ratings in MHz in the part number suffix. A 486DX2-66, for example, is rated to run at 66MHz. Some processors incorporate clock multiplication, which means they can run at different speeds compared with the rest of the system.

Most systems that use the 386 or earlier processors are socketed for a math coprocessor as an option, but they do not include a coprocessor as standard equipment. A few systems on the market at that time didn't even have a socket for the coprocessor because of cost and size considerations.

These systems were usually low-cost or portable systems, such as older laptops, the IBM PS/1, and the PCjr. Most often, you can learn which CPU and math coprocessor are installed in a particular system by checking the markings on the chip.