Least recently used replacment method with protection

Abstract

An LRU with protection method is provided that offers substantial performance benefits over traditional LRU replacement methods by providing solutions to common problems with traditional LRU replacement. By dividing a cache entry list into a filter sublist and a reuse list, population and protection processes can be implemented to reduce associativity and capacity displacement. New cache entries are initially stored in the filter list, and the reuse list is populated with entries promoted from the cache list. Eviction from the filter list and reuse list is done by a protection process that evicts a data entry from the filter, reuse, or global cache list. Many variations of protection and eviction processes are discussed herein, along with the benefits each provides in reducing the effect of unwanted displacement problems present in traditional LRU replacement.

Description

Invention field [0001] The present invention generally relates to a replacement method for cached data in an electronic computer system, and more particularly to a least recently used replacement method that can prevent unnecessary replacement. Background technique [0002] Cache memory is often used in high-performance computing systems, where fast system speed requires fast, expensive memory to quickly access data. The faster the processor works, the faster it must retrieve data and instructions from memory. This requires that the memory can be accessed quickly, and involves the memory being generally located near the processor due to the high clock speed. However, fast memory that can operate as a cache is relatively expensive, and it is often not feasible to place a large or customized amount of memory close to the processor. Therefore, a limited number of cache memory banks are often isolated from the larger main system memory and placed near the processo...

AI Technical Summary

Problems solved by technology

Also, the LRU algorithm may replace frequently used data with recently used data, even if recently used data is used less frequently or only once

Images