Difference between Simultaneous and Hierarchical Access Memory Organisations
Last Updated :
23 Sep, 2024
In the Computer System Design, memory organization is important in order to enhance system performance and efficiency . So, two ways have been devised on the basis of which CPU tries to access different levels of Memory. These two types include Simultaneous Access Memory Organization and Hierarchical Access Memory Organization . Let's explore these memory organizations through this article.
What is Simultaneous Access Memory Organization?
In this memory organization, CPU can access the data simultaneously from all levels of memory i.e. if it was trying to access data from level 1 but it didn't get. So, it can check in other levels at the same time as shown in the figure below.
Figure -Simultaneous Access Memory Organization
Average Memory Access Time for Simultaneous Access Memory Organization
Let there be total three levels in a memory i.e. L1, L2, L3 and H1, H2,H3 are the hit ratios for data to be found in respective levels. T1, T2, T3 be the access time for these levels. Then Average Memory Access time for Simultaneous Access Memory Organization is
Average Memory Access Time
= H1*T1 + (1-H1)*H2*T2 + (1-H1)*(1-H2)*H3*T3
= H1*T1 + (1-H1)*H2*T2 + (1-H1)*(1-H2)*1*T3
= H1*T1 + (1-H1)*H2*T2 + (1-H1)*(1-H2)*T3
Because the chances to get data in L3 will be 100% if it is not found in L2 and L3 , so H3=1 is used here.
What is Hierarchical Access Memory Organization?
In this memory organization, CPU can't access data simultaneously from all levels of memory as it is only connected to level 1 of memory here. So, the data will come from other levels to the level 1 then CPU will get its access as shown in the figure below.
Figure - Hierarchical Access Memory Organization
Average Memory Access Time for Hierarchical Access Memory Organization
Let there be total three levels in a memory i.e. L1, L2, L3 and H1, H2,H3 are the hit ratios for data to be found in respective levels. T1, T2, T3 be the access time for these levels. Then Average Memory Access time for Hierarchical Access Memory Organization is
Average Memory Access Time
= H1*T1 + (1-H1)*H2*(T1+T2) + (1-H1)*(1-H2)*H3*(T1+T2+T3)
= H1*T1 + (1-H1)*H2*(T1+T2) + (1-H1)*(1-H2)*1*(T1+T2+T3)
= H1*T1 + (1-H1)*H2*(T1+T2) + (1-H1)*(1-H2)*(T1+T2+T3)
Again, H3=1 will be used here as it was used above for average memory access time in case of Simultaneous Access Memory Organization.
Also, if data is not found in L1 then time to access data will be T1+T2 while checking for it in L2 and so on for further levels.
Difference between Simultaneous and Hierarchical Access Memory Organizations
| Simultaneous Access Memory Organization | Hierarchical Access Memory Organization |
|---|
| In this organization, CPU is directly connected to all the levels of Memory. | In this organization, CPU is always directly connected to L1 i.e. Level-1 Memory only. |
| CPU accesses the data from all levels of Memory simultaneously. | CPU always accesses the data from Level-1 Memory. |
| For any "miss" encountered in L1 memory, CPU can directly access data from higher memory levels (i.e. L2, L3, .....Ln). | For any "miss" encountered in L1 memory, CPU cannot directly access data from higher memory levels(i.e. L2, L3, .....Ln). First the desired data will be transferred from higher memory levels to L1 memory. Only then it can be accessed by the CPU. |
| If H1 and H2 are the Hit Ratios and T1 and T2 are the access time of L1 and L2 memory levels respectively then the Average Memory Access Time can be calculated as:
T=(H1*T1)+((1-H1)*H2*T2
| If H1 and H2 are the Hit Ratios and T1 and T2 are the access time of L1 and L2 memory levels respectively then the Average Memory Access Time can be calculated as:
T=(H1*T1)+((1-H1)*H2*(T1+T2)
|
NOTE:
- By default the memory structure of Computer Systems is designed with Hierarchical Access Memory Organization. It is so because in this type of memory organization the average access time is reduced due to locality of references.
- Simultaneous access Memory organization is used for the implementation of Write Through Cache.
- In both types of memory organization, the Hit Ratio of last memory level is always 1.
Conclusion
In conclusion, Simultaneous Access Memory Organization is a technique that allows CPU to check for the required data in all levels of memory at the same time. While Hierarchical Access Memory Organization is a technique that allows CPU to access only level 1 of memory for searching the required data. If not found, then it will fetch data from higher levels to level 1 using locality of reference. In this way, both techniques can be used as per their requirement so that system performance can be enhanced.
Related Articles
Simultaneous and Hierarchical Cache Accesses
Locality of reference and Cache operations in Cache memory
Cache Memory Performance
Difference between Temporal Locality and Spatial Locality
Cache Memory - Computer Organization and Architecture
Similar Reads
Simultaneous and Hierarchical Cache Accesses
Prerequisite : Cache Organization Introduction :In this article we will try to understand about Simultaneous Cache access as well as Hierarchical Cache Access in detail and also understand how these access actually works whenever CPU (Central Processing Unit) requests for Main Memory Block which is
9 min read
Difference between Uniform Memory Access (UMA) and Non-uniform Memory Access (NUMA)
In computer architecture, and especially in Multiprocessors systems, memory access models play a critical role that determines performance, scalability, and generally, efficiency of the system. The two shared-memory models most frequently used are UMA and NUMA. This paper deals with these shared-mem
5 min read
Difference between Direct and Immediate Addressing Modes
Whenever an instruction is executed, it requires the operands or data to operate on their own. There are different ways of accessing operands or data. Such methods are known as addressing modes. There are many addressing modes but in this article, we will learn about direct and immediate addressing
3 min read
Difference between Register Mode and Register Indirect Mode
In the computer system, addressing modes refer to how the operands of an instruction is located or accessed. There are two critical addressing modes namely the Register Mode and Register Indirect Mode. These modes are often used to address data present in the registers in the course of running of pr
5 min read
Difference between Random Access Memory (RAM) and Content Addressable Memory (CAM)
In the world of computation, memory plays a very significant role in ensuring efficient data processing and data storage or data retrieval. Memory system are the very crucial part of any particular digital devices , it facilitate to store and retrieve the Data from the devices . Two important types
8 min read
Difference between Cache Memory and Register
In the context of a computer’s architecture, two terms that may be familiar are the cache memory and the register, however, few people may not know the differences and functions of each of the two components. Both of them are important to the CPU but they have different roles and are used in differe
5 min read
Difference between Memory based and Register based Addressing Modes
Prerequisite - Addressing Modes Addressing modes are the operations field specifies the operations which need to be performed. The operation must be executed on some data which is already stored in computer registers or in the memory. The way of choosing operands during program execution is dependen
4 min read
Difference between 2-address instruction and 1-address instructions
When we convert a High-level language into a low-level language so that a computer can understand the program we require a compiler. The compiler converts programming statements into binary instructions. Instructions are nothing but a group of bits that instruct the computer to perform some operatio
4 min read
Difference between 3-address instruction and 2-address instructions
In computer architecture, instructions are the basic building blocks of execution that manually inform records how records is processed through the CPU. Three-operand instructions and two-address instructions differ in how they specify operands for acting operations, such as addition or multiplicati
8 min read
Difference between Memory and Hard Disk
Memory and hard disk are two essential components of a computer system. Memory refers to the temporary storage used by a computer to perform its operations, while a hard disk is a permanent storage device used to store data and programs. 1. Memory: Memories are made up of registers. Memory refers to
4 min read