You are on page 1of 8

Basics of

Computer
Architecture
For 3rd year communication students

DR. ENG. Assem Badr

2014
Basics of Computer Architecture

 Draw the memory hierarchy for the most computers.


Multi-cores Microprocessor

Registers

Capacity Primary cache memory (L1)


SRAM

Secondary cache memory


(L2)

Extra cache memory (L3)

Main memory (RAM)


DRAM

Secondary memory
Speed Magnetic
and (Hard disk) surface
Cost/cell

Auxiliary memory

(Floppy, CR-ROM, DVD)

Revision question 5 Page 1


Basics of Computer Architecture

 What is the meant by SRAM 128KX8 bit?

128K X 8

Location 1
Location 2

FF FF FF FF FF FF FF FF

17
Address bus

Location 131072

Data bus
The left number (128k) has two meanings:-

 Total number of address lines = log 2 (128 k )  17


 Total number of memory locations = 128k  128 1024  131072
The right number (8) has two meanings:-
 Total number of data lines = 8
 Total number of cells (FFs) per single location = 8

 Characteristics for Flash memory


It is permanent memory (nonvolatile memory), it’s modified from electrical
erasable programmable read only memory (EEPROM) with these features:-

 It fabricated with thin layers of semi-conductor.


 Its cell occupies little volume in chip.
 Its package (chip) contain high amount of cells (high capacity).
 It needs single and low voltage supply for reading and writing.
 It consumed low power and dissipated low heat.

Revision question 5 Page 2


Basics of Computer Architecture

 It used in the portable devices such as digital cameras, mobiles…etc.

 Characteristics for Cache memory


It is temporary memory (volatile memory), it’s modified from static memory
(SRAM) with following features:-

 It fabricated with thin layers of semi-conductor with low impurity.


 Its cell (flip-flop) occupies little volume in chip.
 It has high cost cell.
 Its package (chip) contains high amount of cells (high capacity).
 It needs low voltage supply for reading and writing.
 It consumed low power and dissipated low heat.
 It is the closer memory to CPU because it faster memory.
 It used as cache memory level 1 and level 2 (L1 & L2) in CPU and
cache memory level 3 (L3) on mother-board.

 Virtual memory
It isn’t physical memory (imagine memory), it performed by the operating
system (such as windows). It provides by software application “kernel” to
manage large size programs into the main memory (RAM) with low capacity.

For more illustration consider this scenario, assume the current running
programs have 12 GB and the capacity of RAM is 8 GB. It’s clear that the
RAM capacity not enough to load with all running programs simultaneously.

The operating system provides management for the running programs using
the limited capacity RAM as follows:
- It divides the current running programs into arrangement blocks in the
hard disk.
- It loads some of these data blocks (in our case 2 blocks) into the RAM.
- The 1st loaded block is executed by CPU.
- After executed the 1st block it reloads again into hard disk.
- The executed block is replaced with the subsequent block (3rd block) from
the hard disk during the execution of the 2nd loaded block by CPU.
- After the 2nd block is executed it reloads again into the hard disk and
replaced with the subsequent block and so on.
- The operating system (memory management unit) loads the running
program blocks from the hard disk into RAM consecutively, also the loaded
blocks in RAM is delivered to CPU sequentially.

Revision question 5 Page 3


Basics of Computer Architecture

Memory
management
unit (MMU)
Block 1
Block 2 CPU
3
Block 3
1

1
Cache
4 2 memory
Block 1
Block 2

HDD RAM

 Compare between static RAM and dynamic RAM


SRAM DRAM

Each cell is constructed with Each cell is constructed with


more than five transistors capacitor and transistor

Each cell occupy larger volume Each cell occupy smaller


in its chip volume its chip

Its chip contains littel amount of Its chip contains larger amount
cells of cells

Higher cost per cell Lower cost per cell

Don’t needs refresh Needs refresh (recharge


capacitors)

Lower access time (higher higher access time (lower


speed for reading and writing) speed for reading and writing)

Direct addressing through Row/Column addressing

Revision question 5 Page 4


Basics of Computer Architecture

address bus through address bus

Used as cache memory (L1 and Used in main memory (RAM)


L2) because it faster memory because it contains large
amount of cells/chip

Notes

 SDRAM is synchronous dynamic RAM which means that it needs clock


pulses from CPU to read or write.
 DDR SDRAM is double data rate synchronous dynamic RAM which
means that it can operate two reading or/and writing in single clock pulse
from CPU.

 What the different between EPROM and EEPROM?


They are non-volatil memory are designed for storing permanent data.
EPROM EEPROM

Its cell is constructed by field Its cell is constructed by field


effect transistor (FET) with thin effect transistor (FET) with thin
layer of silicon oxide between its layer of silicon oxide between
junctions. its junctions in addition to
electric electrodes is connected
with the silicon oxide layer.

Its cell programs when high Its cell programs when high
voltage (Pp.) is applied on this voltage (Vpp) is applied on this
silicon oxide layer, it becomes silicon oxide layer, it becomes

Revision question 5 Page 5


Basics of Computer Architecture

with negative charge. with negative charge.

Its cell erases when ultra-violet Its cell erases when high
(UV) ray is emitted on this silicon voltage is applied on this silicon
oxide layer, it becomes normal oxide layer, it becomes normal
charge again. charge again.

Ultra-violet ray erases all cells in


Electric voltage erases
approximate interval time 1 to 1.5
particular(s) in few mSecs.
hour.

It needs to remove from its Don’t needs to remove from its


embedded circuit to erase it. embedded circuit to erase it.

Its chip has transparent glass to


Its chip hasn’t transparent glass
pass the UV ray

 What are the Cache memory levels for computer?


The cache memory is nowadays fabricated as static RAM type. Cache
memory is the closest memory to CPU. It stored by the most frequently
instructions and data for CPU. Recently, it divided into 3 categories:
L1 cache … is built inside processor chip. It's a smallest memory capacity,
its range in between 8 KB and 128 KB but it the fastest memory.

Revision question 5 Page 6


Basics of Computer Architecture

L2 cache … is built also inside processor chip. It's slightly larger in capacity
than L1, its range in between 512 KB and 8 MB but it slightly slower than L1
memory.
L3 cache … is built on the motherboard near the processor chip. It's slightly
larger in capacity than L2, its range in between 8 MB and 24 MB but it
slightly slower than L2 memory.

 Draw the cache memory organization for the daul-cores


processor.
Double-cores Microprocessor

Core 1 Core 2

Cache memory L1 Cache memory L1

Cache memory L2

Note: the Core 2 duo processor is a brand name for dual core processor

Revision question 5 Page 7

You might also like