Thursday, January 9, 2014

How To Cache RAM.

How To Cache RAM.

The memory chip random access (RAM) are found on all computers, but not always memory chips
are made the same. Some are faster to renew with electric charges the capacitors that take care of the preservation
the data. The renewal rate - usually expressed in nanoseconds, or one billionth of a second - the speed influences
with which data can be moved from memory to the microprocessor manipulates the data.

The faster the RAM chips are more expensive they will be. To keep the cost of personal computers, various
manufacturers use slower memory chips in the main memory of the PC and some RAM chips faster and
more expensive motherboard, to be used as external cache RAM. The cache - usually between 64 and 256 kilobytes of
Memory - helps move data between main memory and the processor, in the shortest time possible. A RAM cache
has the same effect on the process of speeding up access to a memory disk cache in disk access.

Without cache, the processor could be several clock cycles idle, waiting for the requested data will be
transmitted. A clock cycle is the shortest space of time during which a transaction can last for a computer. To
cache, the computer can keep what you need at your fingertips, ie, the data most likely to be requested by
microprocessor. In the faster chips, data can be supplied to the processor with minimum wait time, the
zero times.

Two factors control the efficiency of cache RAM. One is the speed of the chips used in the cache - the quickest
are, the better. The other factor is the algorithm that uses the cache to determine which extra data will be stored in
Cache. The better the algorithm to discover what data will be required to follow the cache more often mark a
point or hit (a term used to describe every time the processor requests a data cache and can provide it
with their faster chips, instead of having to get it with chips slower main RAM).

When new data is requested by the program that you use, the cache replaces the oldest data that is on the chips
higher speed by the new data and the other data that are around the memory addresses. This obeys the rule
FIFO (acronym for First In, First Out, first in, first out), based on the principle that since there are more
time no use is less likely to be requested in the future by the program.


Source: Evolution of Computers

No comments:

Post a Comment