You are on page 1of 37

ASIA PACIFIC INSTITUTE OF INFORMATION TECHNOLOGY APIIT Diploma Part II

INCOURSE ASSIGNMENT CSA Case Study 3 Memory Architecture


Prepared By A.S.Tiffaniya Fernando (CB004165) A.N.Ahamed Nishadh (CB004081) S.D.Ilangakoon (CB004041) Module Code & Title AICT001-3-2 Computer Systems Architecture Cohort DF10A1ICT Date of Submission 12th May 2011 Instructor Dr.Gayan Jayakody

Submitted in partial fulfillment for the degree of Bachelor of Science (Hons) in Computing

Word Count [2900 words]

Tiffaniya Fernando

Ahamed Nishadh

Deshan Ilangakoon
Assumptions (5) Report content (10) Diagrams and/or Figures (20) Analytical and Critical Thinking (20) Documentation (5) Referencing (10)

GROUP COMPONENT

i
Group Total (70) Presentation (20) Progress Report (5) Workload Matrix (5) Individual Total (30) Grand Total (100)

INDIVIDUAL COMPONENT

WORKLOAD MATRIX
Name / CB No. Work Assigned Work Completed Comments & Signature Tiffaniya Memory Hierarchy Tiffaniya Fernando CB004165 Introduction of the document Proofing the document RISC Architecture Ahamed Nishadh CB004081 Compiling the Document Frequently Asked Questions ALL COMPLETED ALL COMPLETED Ahamed Deshan

Deshan Ilangakoon CB004041

Cache memory GANTT Chart Meeting Minutes ALL COMPLETED

ii

PROJECT GNATT CHART

iii

ACKNOWLEDGEMENTS
Firstly we would like to thank our lecturer Dr.Gayan Jayakody for all the help and guidance given to us while doing this assignment. Especially for teaching us this module in such a way that we were able to learn this highly technical module very easily. Also there are many individuals who have helped us in numerous ways directly and indirectly so that we were able to complete this assignment. APIIT Lanka for providing us with resources and the Tech Team at APIIT Lanka for their assistance at required times. And last but not least our friends, parents and the well-wishers without whose moral support and encouragement, we would not have been able to do a good job. Finally, if there are any shortcomings in this project, then we request to excuse us for all those and accept this documentation. Tiffaniya Fernando Ahamed Nishadh Deshan Ilangakoon

iv

TABLE OF CONTENTS
WORKLOAD MATRIX .............................................................................................. ii PROJECT GNATT CHART ....................................................................................... iii ACKNOWLEDGEMENTS ........................................................................................ iv TABLE OF CONTENTS ..............................................................................................v 1.0 INTRODUCTION ................................................................................................1 2.0 MEMORY HIERARCHY ...................................................................................2 2.1 PRINCIPLE AND BENEFITS ........................................................................ 2 2.2 HOW CAN THE MEMORY HIERARCHY PROVIDES BOTH FAST AND THE LARGE CAPACITY ....................................................................................... 3 2.3 WHAT TYPICAL PROGRAM BEHAVIOR COINCIDES WITH THE BENEFITS OF A MEMORY HIERARCHY? ........................................................ 4 3.0 CACHE ................................................................................................................6 3.1 CACHE MEMORY WRITE STRATEGIES .................................................. 7 4.0 REDUCED INSTRUCTION SET COMPUTERS ..............................................9 4.1 ADVANTAGES OF USING REGISTERS OVER MEMORY ................... 10 5.0 FREQUENTLY ASKED QUESTIONS ............................................................12 6.0 CRITICAL EVALUATION ..............................................................................16 6.1 TIFFANIYA FERNANDO ........................................................................... 16 6.2 AHAMED NISHADH................................................................................... 18 6.3 DESHAN ILANGAKOON ........................................................................... 20

BIBLIOGRAPHY .......................................................................................................22 TABLE OF FIGURES ................................................................................................24 APPENDIX .................................................................................................................25 1.0 MEETING MINUTES .................................................................................. 25 1.1 1st MEETING ............................................................................................. 25 1.2 2nd MEETING ............................................................................................ 27 1.2 3rd MEETING ............................................................................................ 28 1.2 4th MEETING ............................................................................................ 29 1.2 5th MEETING ............................................................................................ 30

vi

1.0 INTRODUCTION
The Computer System Architecture assignment that was handed out on the 10th March 2011 is a Research based assignment. It was separated into 3 Case Studies of which one has to be selected to research upon. The Case Study that our group decided upon to carry out the research on is the 3rd Case Study Memory Architecture. This Case Study contained 3 parts. The first relating to Memory Hierarchy, the second Cache Memory and the third is RISC Architecture. This assignment was carried out by our group and the complied document containing the result of our research is given in this documentation

2.0 MEMORY HIERARCHY


2.1 PRINCIPLE AND BENEFITS
Memory hierarchy is designed is based on three important principles: Make the common case fast. Principle of locality Smaller s faster

A memory system is a hierarchy of storage devices with different capacities, costs, and access times. A memory hierarchy is successful if the data and instructions are available, when needed. On the top of the hierarchy. CPU registers hold the most frequently used data. Small, fast cache memories nearby the CPU act as staging areas for a subset of the data and instructions stored in the relatively slow main memory. The main memory stages data stored on large, slow disks, which in turn often serve as staging areas for data stored on the disks or tapes of other machines connected by networks. A memory hierarchy is successful if data and instructions are available, when needed, on the top levels of the hierarchy. (Computer Systems: A Programmer's Perspective, 2011) The levels of memory in a computer. From fastest to slowest speed, they are shown in the figure shown below:

Figure 1 - Memory Hierarchy A memory hierarchy is computer storage established at each level in the hierarchy by response time. Since response time, complexity and capacity are related. The levels may be established by controlling technology. There are four major storage levels: Internal Processor registers and cache. Main the system RAM and controller cards. On- line mass storage Secondary storage. Off-line bulk storage Tertiary and Off-line storage.

This is a most general memory hierarchy structuring. Many other structures are useful. For example, a paging algorithm may be considered as a level for virtual memory when designing computer architecture. (www.answers.com, n.d.)

2.2 HOW CAN THE MEMORY HIERARCHY PROVIDES BOTH FAST AND THE LARGE CAPACITY
These are the levels in typical memory hierarchy. Moving farther way form CPU the memory in level become larger and slower.

Figure 2 - Memory Path The above principles suggest that we should try to keep recently accessed items in the fastest memory. Because the smaller memories are more expensive and faster, we want to use smaller memories to try to hold the most recently accessed items close to the CPU and successively larger (and slower, and less expensive) memories as we move away from the CPU. This type of organization is called a memory hierarchy. Two important levels of the memory hierarchy are the cache and virtual memory. According to the principles of the memory hierarchy, smaller piece of hardware will generally be faster than larger pieces. There are methods to improve the memory accesses: Wide path Memory Access Retrieve multiple bytes instead of 1 byte at time. Memory Interleaving Partition memory into subsections, each with its own address register and data register Cache Memory

2.3 WHAT TYPICAL PROGRAM BEHAVIOR COINCIDES WITH THE BENEFITS OF A MEMORY HIERARCHY?
As discussed in this work, objects typically are small, short-lived and frequently accessed during their short life span. Thus object references exhibit certain characteristic properties that make the object fields referred by them suitable candidates to be cached separately in a cache structure known as an Object-cache. An Object-cache is a cache structure, separately designed for caching only objects. It
4

creates a parallel data access path for objects. The object references are directed through the Object-cache while the other references are directed through a regular data cache. The performance and energy behavior of the benchmark programs was found to be extremely sensitive to the configuration of the Object-cache, when the optimal configuration was determined on a per-application basis. The relationship between the configuration and performance and energy characteristics of a program (of the entire application) can be explained as follows. A direct mapped Object-cache is more energy efficient per access, consuming only about 30% of the energy of a similar sized four-way set associative cache. This reduction occurs because a direct mapped cache accesses only one tag and data array per access, while a four-way cache accesses four tag and data arrays per access. A direct mapped cache can also have a shorter access time in part because multiple data arrays need not be multiplexed. While a direct-mapped caches hit rate may be acceptable for many applications, for some applications a direct-mapped cache exhibits a very poor hit rate and hence suffers from poor performance and energy consumption.

3.0 CACHE
Cache memory is a type of memory subsystem that is used to reduce the CPUs accessing time of memory. The CPU has an average execution time of approximately two nanoseconds and the accessing time of the RAM is near 60 nanoseconds. This means that the CPU will spend more time looking for the next instruction than actually executing it. This is where Cache Memory comes into use. While cache memory is a lot smaller in size when compared with RAM and other hard drives it is almost double in speed when compared with RAM but it is also very expensive and therefore is used only in small amounts. (Gache, 2008) Cache Memory uses the logic that memory that is accessed is not at random. Memory does not get placed all over but memory of one program is generally near another or in a localized area. There are two types of locality: 1. Temporal Locality: These are data or instructions that the CPU will have to access many times during the execution of a program. 2. Spatial Locality: This is data or instructions that come one after the other in terms of memory addresses. For example arrays. (Jayakody, 2011) Cache can divided into 2 levels. Level 1 (L1) is small in size with about 4-16KB capacity and has an access time of approximately 10 nanoseconds. This cache level is built right into the microprocessor. The next is Level 2 (L2). This cache level has a higher capacity of up to 512 KB with a slightly slower accessing time of 20 to 30 nanoseconds. However CPUs cannot have only L1 cache memory as this would be expensive and also the CPU does not require such large amount of cache due to the above mentioned locality of reference or addresses. Cache only has to load the data or instructions that will be needed immediately and the CPU can then access this data or instructions much faster.

As mentioned above cache is used to reduce the accessing time of the CPU. How this works is the CPU will initially look for the required data or instructions in the cache memory. If it is available then the required data or instructions are immediately loaded into the processor. This is called as a Cache Hit where the required data or instructions are found in the cache memory. However on certain occasions the required instruction or data may not be available in the cache memory and the CPU is then forced to look for it in the RAM. When a situation like this occurs it is called a Cache Miss. When this happens the CPU wastes valuable processor time looking for this data in the RAM. (Gache, 2008) The usefulness of Cache memory is determined by calculating the Hit Rate. This is calculated by dividing the Cache Hits by the sum of the Cache Hits and Cache Misses and the result is given as a percentage.

For a cache memory to be effective the Hit Rate should be greater than 90%. If lower than this the Cache memory is not useful.

3.1 CACHE MEMORY WRITE STRATEGIES


There are three main write strategies that are used to write to cache memory: 1. Write-back 2. Write-through 3. Buffered Write Write-back also known as Copy-back is when a write is made to the current cache memory. However this write is made only to the cache memory and not to the system memory. This will stay in the cache till a new system address requires that space in which instance the data is written into the system memory and the space taken up is made available. The main advantage to Write-back or the copy-back policy is that it reduces the write load to the memory. However this method slows

down when it comes to loading new cache lines as it has to first remove the existing line, write it to memory and then only replace the cache with new memory. Write-through the write is made to the cache and the related memory at the same time. The main advantage of Write-back policy is that both cache and memory copies get updated at the same time. This is especially useful in multiprocessor systems. However this method has a drawback, which is that it tends to waste bus and memory bandwidth. Buffer write on the other hand does not write to memory immediately. However it buffers the write and writes at a later instance and is useful when using a write through policies. (www.pcguide.com, 2001) and (Dandamudi, 2003)

4.0 REDUCED INSTRUCTION SET COMPUTERS


The RISC architecture is a relatively new type of CPU architecture that is very much different in how it works compared to the CISC CPU architecture. As the name itself suggests, this architecture is where certain functionality of the CPU is changed to improve the efficiency and performance of the CPU. And thus to improve the performance of the CPU, designers researched on and found out that certain instructions are used very often than some others and also that most programmers avoid using complex instruction sets in programs. And therefore to optimize the performance, these unwanted wastages were cut down to increase the performance. Main features of RISC architecture are: 1. Limited set of simple instructions which can be executed in a very short time using the optimized CPUs. 2. Register oriented instruction with limited access to memory so that the instructions are executed faster. 3. Its a length independent architecture which means that the length of the instructions is fixed and it doesnt require calculation of previous instructions to continue with the next which again makes the execution of instructions faster. (Jayakody, 2004) Some examples of processors using the RISC architecture are the SPARC (Scalable Processor ARChitecture) by Sun Microsystems, MIPS (Microprocessor without Interlocked Pipeline Stages) by MIPS Technologies, IBM R6000, DEC Alpha by Digital Equipment Corporation and POWER-PC (Performance Optimization With Enhanced RISC Performance Computing) by Apple, IBM and Motorola together.

In the current era, it can be seen that the Apple iPods, Apple iPhones, Nintendo Gameboy devices, Sony Walkman devices and some mobile phones also using CPUs with the RISC architecture implemented into them. (Cook, 2008)

4.1 ADVANTAGES OF USING REGISTERS OVER MEMORY


The main advantage of using registers over memory in instructions is that registers work faster than normal memory. It can be seen in the memory hierarchy diagram in section X above that registers are faster than normal memory modules in terms of data store and load. Thus when using registers, the instruction can be executed faster as very few instructions access the memory thus the time wasted in retrieving is reduced. Also the number of clock cycles per taken for each instruction to be executed is reduced when registers are used. When the memory is being accessed, the number of clock cycles taken is more compared to the cycles taken when accessing the register.

Figure 3 - CISC v RISC

10

When the numbers of clock cycles are less and the instructions get executed faster, the performance of the computer is increased. More work can be done in less time thus making the computer more efficient. One more advantage in this situation is the cost involved. Although memory is cheaper than registers, in the CPU designs, the registers cost less than transistors which are used mainly in the CISC computers. And so the cost of the CPU is also brought down and the concept of Very Large Scale Integration is brought into work at this instance as more space is available when transistors are not used hence more registers can be used or the size of the CPU can be reduced. (Wang & Horvath, n.d.) Also in the case of RISC architecture, its more favored by the programmers. It is a known fact that very few programmers tend to write complex instruction sets. Programmers often right simple instruction sets that are faster to execute and so the RISC architecture helps them in their work. (Swanson, 2003)

11

5.0 FREQUENTLY ASKED QUESTIONS


Q : What is Memory ? A : Memory can be referred to the components in a computer or electronic device where data can be stored for running the device. Memory are of different types and are used according to the need. Q : What is a Memory Hierarchy ? A : A memory hierarchy is the hierarchy of memory. It is usually ordered in the speed of the memory type. For example, Registers are at top as they are very fast while hard disks are at last as they are comparatively slower. Q : What are the different types of memory ? A : Registers, Cache Memory, Main Memory, Magnetic Disk, Magnetic Tapes and other disks. Q : What are the 4 storage levels ? A : Internal, Main, On- line mass storage, Off-line bulk storage Q : What is Cache Memory? A : Cache memory is used by the CPU to reduce the speed it takes for reading memory by keeping back-ups of the most frequently and recently used data and instructions. Q : How is Cache Memory used ? A : When the CPU needs to Read or Write data to the main memory it Reads and Writes to the Cache since it is a lot faster than Reading or Writing to the main memory. This helps the CPU to increase its speed since main memory takes a longer period to execute than the CPU Q : What is Locality?

12

A : Locality is the Logic that is used behind Cache memory. This logic says that the memory addresses of the memory that the CPU needs to access is not at random. Instead it is either in a temporal, this refers to repetitive memory, or spatial, which refers to memory that comes in a sequential manner. Q : What are write strategies used for? A : Write strategies determine how the data or instructions that needs to be written into the main memory and the cache is done. There are three methods or policies. They are Write-back, Write-through and Buffered Write. Q : What are Cache Hits and Misses? A : When the CPU looks for the data or instructions that it needs in the Cache and finds it, it is known as a Cache Hit. If it does not find the data or instructions in the Cache it is known as a Cache Miss. Q : What is a CPU ? A : The CPU, also known as the processor of a computer is the main element of the computer system. It is where all the calculations take place to run the computer. Q : Name two types of CPU Architectures ? A : There are many CPU architectures that are used when designing CPUs. Two of the main architectures are the Complex Instruction Set Computer Architecture and the Reduced Instruction Set Computer Architecture. Q : What is a Cycle in terms of CPU Cycle ? A : A CPU Cycle refers to a single pulse of the computer clock. For example, a 4 MHz CPU will have 4 million CPU cycles per second. (http://www.codepedia.com, 2004) Q : What is Very Large Scale Integration ? A : VLSI is an integrated circuit design that contains 5,000 to 50,000 components on a single chip. (http://www.computerhope.com, n.d.)

13

Q : What is a RISC Architecture ? A : Developed and introduced by IBM in 1980, RISC is short for Reduced Instruction Set Computer and pronounced as risk. RISC is a processor architecture that requires fewer instructions to operate, causing the processors to be faster than earlier CISC processors. (www.computerhope.com, n.d.) Q : Examples for RISC Architecture ? A : SPARC (Scalable Processor ARChitecture) by Sun Microsystems, MIPS (Microprocessor without Interlocked Pipeline Stages) by MIPS Technologies, IBM R6000, DEC Alpha by Digital Equipment Corporation and POWER-PC (Performance Optimization With Enhanced RISC Performance Computing) by Apple, IBM and Motorola together. Q : How many levels are there in cache memory? A : Two levels, Level1 and Level2. Q : Which level is faster? A : Level1. Q : Which level has more memory capacity? A : Level 2 Q : What is Write -Back? A : Write-Back writes to cache only not the main memory. Q : What is Right-Through? A : Write through to both cache and main memory. Q : What is Buffered- Right? A : Buffers write instruction until a later moment. Q : What is the Hit rate?

14

A: Q : What are the types of locality? A : Temporal Locality and Spatial Locality. Q : What is Temporal Locality? A : Memory that it used repetitively. Q : What is Spatial Locality? A : memory that is accessed in sequence order. Q : In memory what is the speed to capacity relationship? A : Greater the capacity, Slow the speed. Q : Whats the average access time of cache? A : 10 nanoseconds. Q : What are the main two types of RAM? A : SDRAM and DRAM. Q : What are the two main types of ROM? A : PROM and EPROM Q : What are the main major storage levels? A : Internal, main, off-line mass and off-line bulk. Q : What are the two important levels of memory hierarchy? A : Cache and Virtual Memory. Q : What other types of CPU Architectures are available? A : RISC, CISC, VLIW, etc.

15

6.0 CRITICAL EVALUATION


6.1 TIFFANIYA FERNANDO
My group members and me decided to work on case 3. This case is about Memory Architecture. As the leader of my group I assigned the questions to my group members according to the case. And I decided to take the first question from the case that is about the Memory Hierarchy. This was a new task for me and I selected this topic because its a topic where I was interested in this module. Though I learned about memory hierarchy in class, this assignment helped me a lot to improve my knowledge on this particular subject. I did my research on this project in various ways, such as browsing online, referred books and my lecturer gave me a fair idea on memory hierarchy. With the help of all these researches and lectures my knowledge on memory hierarchy gradually improved. I had no knowledge on the principle of the memory hierarchy, but however I managed to get help from my other group members whenever I get stuck and also some of the websites from Google helped me a lot to get fair knowledge on the specific topic. I was able to manage the troubles, and I overcome along that. The opportunity of being a group leader was great experience for me, I learnt to handle group members, and I was able to overcome the problems and we shared information among us. I managed to work with my group members in a pleasant way, and also I have been with the same group members in my previous assignments as well. So it was an awesome experience to work the same people as a group leader. As I mentioned before I was assigned to do question one, but this question had three different parts. Even though I already had an idea on this topic the last part of the question I didnt understand, so I got help from one of my group member, after he explained me I started on my research through Internet and now am familiar with the whole question. I had certain knowledge about memory hierarchy; also I knew the memory hierarchy chart, so I drew it on my document.

16

Another task I was assigned to do is finalizing the document, after getting all the documents from my group members. My group members did meeting minutes and the Gantt chart, and my part was to finalize the documentation. As a conclusion, the project was successfully managed for the submission. And It was so thankful to gain more knowledge about memory hierarchy.

17

6.2 AHAMED NISHADH


In this research based assignment I was faced with the problem of researching every little bit of details from scratch as it was a completely new topic. We all use various computerized electronic devices in our day to day lives but we never for a minutes in our busy days stop to think how each device works, what is exactly happening inside the device, what are the components inside the device, how are they designed etc. But after doing this research, I was able to get a very idea on how the main element of any computerized electronic device work. The Central Processing Unit also referred to as the brain of the computer or any computerized device is actually a very interesting component. It is actually amazing to know how such a small component contains so many elements inside it and does so much work At the beginning of this research I had very little idea of what CPU architectures, addressing modes, registers and memory meant as we hadnt by the time I had started the research studied those. As time went on while studying about these at lectures and by reading up on articles on the internet as well as in books I was able to get a comprehensive idea on what these meant and how they are all connected to each other. In my documentation, I have referenced from various sources. It should be noted that most of my references have been made from lecture notes or handouts given at various higher education institutes. I choose to use this information as sources for my documentation as these are information compiled by learned people in those relevant fields of study and the information is authentic and reliable most of the time. This helps me make sure that I have real and known information in my documentation rather than unknown or unreliable information. In addition to this, I have also reference lectures notes given to us by our lecturer. Certain definitions have been taken from websites that are specialized in providing definitions for technical words. I used this source for this information as they are reliable and are being used and trusted by many users.

18

At the end of the research I am now able to understand completely on how CPUs work and the different kind of architectures that are used in designing CPUs, the way they work, how changes in CPU components change the way they work and how it affects the performance of the computer as a whole where those that I gained knowledge on.

19

6.3 DESHAN ILANGAKOON


The group in which I worked for this assignment chose case study 3 and I was assigned to undertake the second segment of that case study which was pertaining to Cache Memory. This confronted me with a new challenge. Though I had previously learnt about Cache Memory this assignment opened my view to help me understand Cache in a wider perspective. In the compilation of information for this assignment I researched in depth to find information regarding Cache memory. My understanding on Cache has vastly improved during the course of this project. I had no knowledge on certain segments such as Cache Write strategies. This posed a fairly strong challenge to me. However with help from the rest of my team members I was able to overcome these problems. I also had to employ self-learning in order to complete this assignment up to the standards that were expected of me. The opportunity to work with the members I have in my group was also a very enjoyable and an extremely pleasant experience. Working with this group was also very easy since I had worked with them on prior assignments and thus I was familiar with them. The work that I was assigned to do as mentioned above was section 2 of the second case study. In this section the work required me to investigate into Cache Memory, specifically on how it works with regards to speed of execution and fetching instructions. I also had to research on what Cache Hits and Cache Misses are and what relevance they have to Cache Memory I also had to research into the Write Policies employed when writing Cache into memory. The main three policies that I researched into were Write-back or Copyback, Write-through and Write with buffer. For the assignment I had to research into how each policy works compare their advantages and disadvantages. Another task that I was assigned to do was assigned was the maintaining of meeting minutes. This task was fairly easy since I had previously also been tasked with handling the meeting minutes.

20

The main problem that I faced during the assignment is the lack of information regarding the Write policy Write with Buffer. However I was able to after searching the internet find a PDF document with information regarding Write with buffer. In conclusion this assignment, though a challenge, was highly fruitful and helped me immensely in gaining a better understanding of Cache memory and I am thankful for that opportunity.

21

BIBLIOGRAPHY
Computer Systems: A Programmer's Perspective, 2011. Chapter 6 - The Memory Hierarchy. [Online] Available at: http://csapp.cs.cmu.edu/public/ch6-preview.pdf [Accessed 2011]. Cook, C.D., 2008. [Online] Available at: http://www.cs.sjsu.edu/~lee/cs147/RISC%20vs.%20CISC.ppt [Accessed 2011]. Dandamudi, S., 2003. Cache Memory. [Online] Available at: http://www.scs.carleton.ca/sivarama/org_book/org_book_web/slides/chap_1_version s/ch17_1.pdf [Accessed 2011]. Gache, G., 2008. How Cache Memory Works. [Online] Available at: http://news.softpedia.com/news/How-Cache-Memory-Works-83803.shtml [Accessed 2011]. http://www.codepedia.com, 2004. CPU Cycle. [Online] Available at: http://www.codepedia.com/1/CPU+cycles [Accessed 2011]. http://www.computerhope.com, n.d. VLSI. [Online] Available at: http://www.computerhope.com/jargon/v/vlsi.htm [Accessed 2011]. Jayakody, D.G., 2004. RISC and CISC Architecture. Lecture Notes. Asia Pacific Institute of Information Technology. Jayakody, D.G., 2011. Type of memory. Lecture Notes. Asia Pacific Institute of Information Technology. Swanson, W., 2003. The Art of Picking Intel Registers. [Online] Available at: http://www.swansontec.com/sregisters.html [Accessed 2011]. Wang, R. & Horvath, T., n.d. RISC vs. CISC. [Online] Available at: www.cs.virginia.edu/~skadron/cs654/cs654_01/slides/tibor.ppt [Accessed 2011]. www.answers.com, n.d. Memory hierarchy. [Online] Available at: http://www.answers.com/topic/memory-hierarchy [Accessed 2011].

22

www.computerhope.com, n.d. RISC. [Online] Available at: http://www.computerhope.com/jargon/r/risc.htm [Accessed 2011]. www.pcguide.com, 2001. Cache Write Policy and the Dirty Bit. [Online] Available at: http://www.pcguide.com/ref/mbsys/cache/funcWrite-c.html [Accessed 2011].

23

TABLE OF FIGURES
Figure 1 - Memory Hierarchy ...................................................................................... 3 Figure 2 - Memory Path ............................................................................................... 4 Figure 3 - CISC v RISC ............................................................................................. 10

24

APPENDIX
1.0 MEETING MINUTES 1.1 1st MEETING
Date Time of Commencement Time of Conclusion Venue Members Present : Tuesday 15th March 2011 : 04.50 pm : 05.15 pm : APIIT, Union Place, 6th Floor Cafeteria : Tiffaniya Fernando Ahamed Nishadh Deshan Ilangakoon Agenda: Deciding on which case study to be undertaken Maintaining meeting minutes and records

Minutes The meeting commenced with Tiffaniya taking the lead and asking the rest of the group which case study would be best to take up and who would want to tackle each part. Deshan said that both case studies two and three were good and Ahamed suggested that since Memory Architecture was the easiest that this should be the topic that would be followed. Since there were no disagreements on this point it was finalized that the group would take on case study 3. The next matter was who should take on which part of the assignment. After a brief discussion it was decided that Part A of the case study would be taken or by Tiffaniya Part B by Deshan and Part C by Ahamed. After this the next matter that was brought up was the recording and maintaining of the minutes of the meeting that is being held and those to come. Deshan volunteered to take on this task and said he would maintain all the records.

25

After this meeting was adjourned until the 30th of March 2011.

. Group Leader

. Lecturer

26

1.2 2nd MEETING


Date Time of Commencement Time of Conclusion Venue Members Present : Wednesday 30th March 2011 : 03.30 pm : 04.00 pm : APIIT, Union Place, 6th Floor Cafeteria : Tiffaniya Fernando Ahamed Nishadh Deshan Ilangakoon Agenda: Reading of previous meetings minutes Assessment of group progress

Minutes The meeting started with a brief re-cap of the previous meetings minutes. After which Tiffaniya asked each member as to how the progress of their allocated sections was fairing. Both Ahamed and Deshan stated that though they had not started to compile the document they had begun research into their sections. After this problems that were being faced by the members were discussed. since it was still in the early stages of the research there were no problems as yet to be discussed. Meeting was adjourned until the 6th of April 2011. . Group Leader . Lecturer

27

1.2 3rd MEETING


Date Time of Commencement Time of Conclusion Venue Members Present : Tuesday 5th April 2011 : 04.50 pm : 05.15 pm : APIIT, Union Place, 6th Floor Cafeteria : Tiffaniya Fernando Ahamed Nishadh Deshan Ilangakoon

Agenda: Reading previous meeting minutes Assessment of work Problems faced

Minutes The main reason for this meeting to be called was to check on the progress of the work that has so far been done. Here all the work done so far was presented and Tiffaniya checked on the work to see if the work done was accurate and adequate. Tiffaniya told all the members present that the work done so far was in her opinion satisfactory and no further change. Tiffaniya then asked the members present as to whether there were any difficulties that they were facing. Deshan stated that he was having a little difficulty finding information on section B of his allocated part 2. Ahamed stated that he was not facing any difficulties as yet but would let the group know if he was. The meeting was after this, adjourned the 26th of April.

. Group Leader

. Lecturer

28

1.2 4th MEETING


Date Time of Commencement Time of Conclusion Venue Members Present : Tuesday 26th April 2011 : 04.50 pm : 05.15 pm : APIIT, Union Place, 6th Floor Cafeteria : Tiffaniya Fernando Ahamed Nishadh Deshan Ilangakoon Agenda: Reading previous meeting minutes Assessment of work Problems faced

Minutes The meeting commenced by going over the last meeting minutes to see if there were any problems that were carried on from the last meeting. Deshan said that he was no longer having difficulty in finding information and was getting along fine with the work. Ahamed pointed out that he was having a few problems as he did not have any prior knowledge of the topic but said he could manage. After this the meeting was adjourned until the 10th of May where all the separate segments would be complied.

. Group Leader

. Lecturer

29

1.2 5th MEETING


Date Time of Commencement Time of Conclusion Venue Members Present : Tuesday 10th May 2011 : 04.50 pm : 05.15 pm : APIIT, Union Place, 6th Floor Cafeteria : Tiffaniya Fernando Ahamed Nishadh Deshan Ilangakoon Agenda: Collect all information from all members and to compile it

Minutes The main purpose of this meeting was so that all the information and documents of each separate member could be collected and compiled. All the members handed over their documents to Ahamed who agreed to compile it and to format it into the specified way. After this Tiffaniya thanked all for the hard work and the meeting was adjourned.

. Group Leader

. Lecturer

30

You might also like