Computer Frequently used data Important data, but it is used not often
Why it was developed? ? ? In 1980 -s the gap in case of performance between microprocessors and memory was increasing. Microprocessors was becoming more powerful and the memory fall behind=> the performance was degrade because of more slower memory than processor. Then it was chosen as a more economical way to use excess amount of low speed memory, but to enter the small but fast cache memory, to mitigate the productivity gap. The result: large amounts of memory, for about the same price and with little loss of overall performance.
• How does it work? • Levels of cache RAM
Types of misses The miss when reading from instruction cache. Usually gives a very high latency, because the processor cannot continue execution of the program and forced to stand in waiting for the downloading of the manual from memory. The miss for reading from the data cache. Usually gives less delay, because instructions not dependent on the requested data can continue to be fulfilled until the request is processed in main memory. After receiving the data from memory to continue the execution of dependent instructions. A miss in the data cache.
Category of misses Compulsory misses (it is caused because of the first reference to the requested address. ) Capacity misses (caused due to the limited size of the cache) Conflict misses (are those misses caused by the conflict. They can be avoided if the cache is not replaced the earlier entry. )
Caching algorithms When the cache is full, the algorithm needs to choose what exactly to remove from it to be able to write (in cash) new, more relevant information. • Most Recently Used • Least recently used • Least-Frequently Used
Why don’t we use Cache instead of Ram? + Extra fast + Very small - Very expensive - Small capacity
The end
Скачать презентацию Computer Frequently used data Important data but it
Cashe Memory.pptx