DTK Apex 386/33, User Manual

Page: 143 / 171

Cache Organization — Direct-Mapped Cache
The direct-mapped cache memory is an alternative to associative-
cache memory, which uses a single address comparator for the
memory system and standard RAM cells for the address and data
cells. The direct-mapped cache is based on an idea borrowed from
software called hash coding.
This is a method for simulating an associative memory. In the hash
coding approach, the memory address space is divided into a number
of sets of words with the goal of each set having no more than one word
of most-frequently-used data.
Each direct-mapped cache address has two parts. The first part, called
the cache index field, contains enough bits to specify a block location
within the cache. The second field, called the tag field, contains enough
bits to distinguish one block from other blocks that may be stored at a
particular location.
For example, consider a 64KB direct-mapped cache that contains 16K
32-bit locations and caches 16MB of main memory. The cache index
field must include 14 bits to select one 16K block in the cache, plus 2
bits to select a byte from the 4-byte sub-block. The tag field must be 8
bits wide to identify one of the 256 blocks that can occupy the selected
cache location. Therefore, the system requires 64KB of cache RAM
(16K 4-byte sub-blocks) to hold the data and code and 16K of 8 bit
RAM to hold the tag. The direct-mapped cache organization is shown
as follows.
24
Chapter 6: Appendix