486 Processor
      Show Part Numbers

The 486 processor contains new architectural features that Intel claims makes it roughly twice as fast as the 386 processors.

Integral math coprocessor
Instruction pipelining
32-bit internal address and external data paths
Floating point unit 386 versus 486
8 kilobyte (KB) internal memory cache
Streamlined hardware design
Tighter silicone design rules

Tighter silicone design rules

The 486 processor pioneered one micron design rules. This allows for the finest of details to be etched into the chip while occupying an area of no more than one micron (one millionth of a meter). The 50 megahertz (Mhz) and 60 Mhz members of the 486 family of processors use 0.8 micron technology. This allows for more transistors to be placed on the chip yet requires less energy and generates less heat than other 486 processors.

Streamlined hardware design

The streamlined hardware design, particularly the pipelining, allows the 486 to work faster than a 386 operating at the same clock speed. The streamlining, such as pipelining and smaller design rules, makes processing of certain multiple clock cycle instructions be completed in one clock cycle.

8 KB cache

For caching read operations, all 486 processors use direct-mapped, write-through cache with one 16-bit line. For simple 16-bit programmed software like MS-DOS and Windows 3.1, this type of cache is sufficient. With more complicated software, a direct-mapped cache can cause thrashing. Thrashing occurs when there are multiple memory requests for the same block of RAM. This was resolved with the Pentium 16 KB cache (double-8 KB design).

This cache holds data and instruction readily accessible for the processor while the processor completes other instructions, thereby reducing the amount of time for the processor between instructions and calculations. The cache then sends the results to the RAM.

The Pentium processor resolves this problem by employing 16 KB (double-8 KB design) of cache. The cache holds data and instructions that are readily accessible to the processor, consequently reducing the amount of time between performing instructions and calculations.

The 486 processor has its cache organized with a four-way, set-associative design. This essentially splits the cache into four 2 KB caches, which enhance its performance with multi-threaded applications.

External caches improve performance in writing to the computer's memory, such as the 64 KB and 256 KB cache chips on the DXLB.

Floating point unit

Intel has added a floating point unit (FPU) to the 486 as compared to the 386 processor, increasing the number of math-type calculations that the processor can perform at any specific time. The FPU calculates any non-integer math calculation and puts the result in the data cache. The floating point unit can rapidly perform non-integer math calculations. This is useful for things such as plotting positions of individual points in a CAD drawing.

32-bit external data path and internal address path

The 486 processor has a 32-bit internal address bus and 32-bit data bus, which means data is processed as fast as the external data bus can carry it. The data bus delivery was enhanced with computers that used 32-bit data buses like extended industry standard architecture (EISA) and VLB technologies, because data is transferred in 32 bits versus 16 bits found on ISA motherboards.

Instruction pipelining

Pipelining allows a microprocessor to read an instruction, start to process it, and before finishing with the instruction, read another instruction.

Integral math co-processor

SX processors do not have integrated math co-processors, but the DX processors have them. The co-processor runs at the same speed as the microprocessor and does not require data to pass through system circuitry. The co-processor increases the performance of the computer when performing mathematical calculations; it also raises the cost.