2. MEMORY CHARACTERISTICS
LOCATION : The memory can either be stored externally with the
help of some devices or internally.
Examples of primary memory devices are internal devices and
secondary memory devices are external devices.
Based on its physical location, memory is classified into 3 types :
On-Chip Memory : This memory is present inside the CPU.
Eg. Internal Registers and L1 Cache.
Internal Memory : The memory is present on the motherboard.
Eg. RAM, ROM
3. .
External Memory : This memory is connected to the
motherboard.
Eg. Hardisk, Pen drive
STORAGE CAPACITY : The amount of data a device can store
is called capacity.
It can be varied for both internal and external devices.
It is measured as a byte (1 byte = 8 bits, 1 bit is either 0 or 1).
This indicates the amount of data stored in the memory.
4. .
It is represented as N x M.
N = number of memory locations ( no of words )
M = number of bits per memory location ( word size )
Eg. (4K x 8) means there are 4K location of 8-bits each.
TRANSFER MODES : The measure of data is different in internal
and external devices.
In internal devices, it is measured as the number of electrical
signals that are moving in and out of the device.
In external devices, it is the number of bits moved inside and out.
5. .
Data can be accessed from memory into 2 different ways :
Word Transfer :
Here, if CPU needs some data, it will transfer only that
amount of data.
Eg. Data accessed from L1 Cache
Block Transfer :
Here, if CPU needs some data, it will transfer an entire
block containing that data.
This makes further access to remaining data of this block
much faster.
This is based on principle of Spatial Locality.
A processor is most likely to access data near the current
location being accessed.
6. .
Eg. On a cache miss, processor goes to
main memory and copies a block
containing that data.
ACCESS MODE : The way of searching the storage devices is
called the access method.
Memories can allow data to be accessed in 2 different ways :
Sequential Access :
Here locations are accessed one by one in sequential
manner.
The access time depends on how far the target location
is, from the current location. Farther the location,
more will be its access time.
Eg. Magnetic Tapes
7. .
Random Access :
Here all locations can be directly accessed in any random
order.
This means all locations have the same access
time irrespective of their address.
Eg. Most modern memories like RAM
PHYSICAL PROTPERTIES : There are various physical
attributes to memory.
Writeable : Contents of the memory can be altered.
Eg. RAM
Non-Writeable : Contents of the memory cannot be altered.
Eg. ROM
Volatile : Content of the memory are lost when power is switched off.
Eg. RAM
8. .
Non-Volatile : Contents of the memory are retained when power is
switched off.
ACCESS TIME (tA) OR LATENCY :It is the time taken between
placing the request and completing the data transfer.
It should be as less as possible.
It is also known as latency.
RELIABILITY : It is the time for which the memory is expected to
hold the data without any errors.
It is measured as MTTF : Mean Time To Failure.
It should be as high as possible.
9. .
COST : This indicates the cost of sorting data in the memory.
It is expressed as cost/bit.
It must be as low as possible.
AVERAGE COST : It is the total cost per bit, for the entire
memory storage.
Consider a system having two memories M1 (RAM) and M2
(ROM).If C1 is the cost of memory M1 of size S1 & C2 is the cost
of memory M2 of size S2.Then the average cost of the memory is
calculated as :
Average cost = C1S1+C2S2 / S1+S2
Small sizes of expensive memory and large size of cheaper
memory lowers the average cost.
10. .
HIT RATIO (H) : Consider two memories M1 and M2. M1 is
closer to the processor (Eg. RAM), than M2 (Eg. Hardisk).
If the desired data is found in M1, then it is called a Hit, else it is a
Miss.
Let N1 be the number of Hits and N2 the number of Misses.
The Hit ratio H is defined as a number of hits divided by total
attempts.
H = N1 / N1+N2
TRANSFER RATE : This is the rate at which data can be
transferred into or out of a memory unit.
For random-access memory, it is equal to 1/(cycle time).
For non-random-access memory, the following relationship holds
:
11. .
tN = tA + n/R
tN = average time to read or write n bits
tA = average access time
n = number of bits
R = transfer rate, in bits per second
13. .
The purpose of any memory device is to store programs and
data.
Several types of memory devices are used in the computer
forming a memory hierarchy.
Each plays a specific role contributing to the speed, cost
effectiveness, etc.
14. .
REGISTERS :
Registers are present inside the processor.
They are basically a set of flip flops.
They store data and addresses and can directly take part in
arithmetic and logic operation.
They are very small in size typically just a few bytes.
PRIMARY MEMORY :
It is the original form of memory also called as main memory.
It comprises of RAM and ROM; both are semi-conductor
memories.(chip memories)
ROM is non-volatile.
It is used in storing permanent information like the BIOS
program.
15. .
It is typically of 2MB - 4MB in size.
RAM is writeable and hence is used for day to day operations.
Every file that we access from secondary memory, is loaded
into RAM.
To provide large amount of working space RAM is
typically 2GB - 8GB.
SECONDARY MEMORY :
The main purpose of secondary memory is to increase the
storage capacity, at low cost.
Its biggest component is the Hard disk.
It is writeable as well as non-volatile.
Typical size of HD is 1TB.
16. .
PORTABLE SECONDARY MEMORY :
These are required to physically transfer files between
computers.
Eg. CD – It is an optical form of storage. Typical size is 700MB.
DVD – It is an optical form of storage. Typical size 4.7MB.
Pen drives and Memory cards – It is a semi-conductor form
of storage.
CACHE MEMORIES :
It is fastest form of memory as it uses SRAM (static RAM).
The main memory uses DRAM (dynamic RAM).
SRAM uesd flip-flops. Hence is much faster than DRAM which
uses capacitors.
17. .
But SRAM is also very expensive as compared to DRAM.
Typical size of cache is around 2MB – 8MB.
If code and data are in the same cache then it is unified cache
else it is called split cache.
Cache memories are the fastest memory but also the costliest.