This essay has been submitted by a student. This is not an example of the work written by our professional essay writers.
Computer Systems Architecture
ACPU cacheis acacheused by thecentral processing unitof acomputerto reduce the average time to accessmemory. The cache is a smaller, faster memory which stores copies of the data from the most frequently usedmain memorylocations. As long as most memory accesses are cached memory locations, the averagelatencyof memory accesses will be closer to the cache latency than to the latency of main memory.
(i) Reasons for cache
The reason for cache is to understand how cache works in a modern day PC and realizing how the basic mechanisms associated to cache may be applied to distribute computing environments. A buffer that holds data that may be accessed, such data is commonly held elsewhere in the system in another buffer. However, this buffer is slower than the cache buffer due to geographic constraints or implementation constraints.
Buffer improve performance, it is to access data via cache that it would be quicker than to access data when the cache is not present. There are some problem with cache which is data that in the cache should be updated frequently and what data we must put in the cache.
(ii) how cache works
The CPU retrieves instruction from memory, acts out the instruction, then retrieves the next instruction (fetch execute cycle). These instructions are very “fine grained”. The CPU may carry out this cycle many times per second. However, the problem with the CPU is that it is very fast, yet accessing instructions from memory is very slow. This is due to geographic location of memory, and implementation type of main memory. The solution is to provide instructions and data that the CPU requires in a buffer that is geographically closer and works at a faster rate than regular memory
(iii) levels of cache
Levels of cache (1 & 2)
Level 1 - Level 1 or primary cache is the fastest memory on the PC building directly into the processor itself. At Level 2 - Level 2 cache is usually found either on the motherboard or a daughterboard that inserts into the motherboard. Pentium Pro processors actually have the level 2 cache in the same package as the processor itself (though it isn't in the same circuit where the processor and level 1 cache reside). Level 2 cache is also sometimes called "external" cache since it resides outside the processor.
Main Memory (level 3)
Cache level 3 is a High-performance processors, such as the Pentium, are fueling the demand for cache RAM, particularly synchronous SRAMs. Static Random Access Memory (SRAM) is a type of memory that is faster and more reliable than the more common DRAM (Dynamic RAM). The term static is derived from the fact that it doesn't need to be refreshed like Dynamic RAM. DRAM supports access times of about 60 nanoseconds while SRAM can give access times as low as 10 nanoseconds. Nevertheless, SRAM is far more expensive than DRAM as SRAM tends to be used for cache only.
(iv) cache memory organization
Caches are introduced into a system to buffer the mismatch between main memory and processor speeds. A cache is a relatively small, fast memory placed between the processor and the main memory. The cache is designed so that its access time matches the processor cycle time. Thus, if the processor is running with a 100MHz clock the cache should be able to respond to a memory request in approximately 10ns. In the high-performance single-chip processors being built today, the cache memory is actually built on the processor chip and separated into distinct instruction and data caches. The typical size of these caches is 8kb, for a total of 16kb of cache on the processor chip.
B (v) write strategies for cache memories
Level 1 is the fastest memory on the PC. It is in fact built directly into the processor itself while Level 2 cache is usually found on the motherboard. Pentium Pro processors actually have the level 2 cache in the same package as the processor itself. Cache level 3 is a High-performance processors, such as the Pentium, are fueling the demand for cache RAM, particularly synchronous SRAMs. Static Random Access Memory (SRAM) is a type of memory that is faster and more reliable than the more common DRAM (Dynamic RAM). Nonetheless, SRAM is far more expensive than DRAM. SRAM tends to be used for cache only.