Memory hierarchy

In computer science, memory hierarchy refers to the arrangement of storage in a computer architecture from the perspective of the main processor, ordered by decreasing access speed, falling costs and increasing capacity. The graphical representation of this hierarchy is usually instead of triangular pyramid shape, and the memory is referred to.

A ( utopian ) ideal memory would be extremely cost-effective, infinite, no access time would need and would also not lose its contents when power is not turned ( non- volatile memory). Would be such a memory is available, the memory management would be trivial. In reality, however, fast memory are orders of magnitude more expensive than slower memory, in addition they are usually volatile. For this reason, an attempt is made to find a compromise between speed, cost and persistence by the memory is created in a memory hierarchy and tries to use by a clever memory management the benefits of the various components, while at the same time circumvent the disadvantages.

This results in the form of a memory pyramid: is located at the tip very little faster and more expensive memory, at the bottom, however, much slower and very low memory. Therefore, while decreasing the rate of top to bottom, increases approximately to the same extent the memory capacity.

A memory hierarchy used on many computers today is:

In addition to the high cost is another reason why very fast memory can be used is limited, the following: The access speed and storage capacity can be further influenced in the higher levels by the spatial arrangement. The main processor is space limited, and there is there a relatively high cooling requirements. The arrangement of the storage groups in height ( ie, the use of the third dimension ) can therefore take place only within narrow limits.

( Eg optimization of the access time for frequently used data) are often used to improve the organization of the memory shifted memory contents within the hierarchy transparent, eg parts of the disk are loaded into the disk cache, on the other hand, parts of memory are swapped out to the pagefile ( virtual memory).

Memory hierarchies in databases

The various memory planes a data are sometimes referred to according to their speed as the primary memory, secondary memory or tertiary memory. It is meant as a primary storage is usually the memory while secondary storage memory and a fast hard tertiary storage means a slow archiving medium.

Hierarchical Storage Management and Information Lifecycle Management

Hierarchical storage management systems store more unnecessary data automatically to a lower memory hierarchy from. Also, information lifecycle management automates such processes.