PHY 421:ELECTRONIC CIRCUITS

AND MICROPROCESSORS

COMPUTER MEMORIES
Introduction

 As the word implies “memory” means the place where we have to store
any thing, this is very essential part of human being just like this memory is
also very important for computer system because in computer system we
have to store some data or information and for storing these items we need
some memory or space.
 memory is very important part of computer system
 There are various units which are used to measure computer memory

 Bit - Smallest unit of computer memory

 Byte - 8 bit = 1 byte

 Kilobyte - 1024 byte = 1 kb

 Megabyte - 1024 kb = 1 mb

 Gigabyte - 1024 mb = 1 gb

 Terabyte - 1024 gb = 1 tb
 Primary Memory :-

 Primary memory also known as “main memory "or “internal memory” which
is located in the mother board of system or as we say which is directly
connected to the CPU. It is the place where only little bit of data are stored
either by manufacturer or by user.

 This is further divided into two parts :-

 RAM

 ROM
 Secondary memory
 The place where we store our personal data in computer system is known as
SM. In this we store data so that when the data is needed we can easily retrieve
it.
 This is the fundamental component of comp. system.
 It is non volatile in nature so that we can not loose the data when power supply
is off.
 There are two methods for accessing the data from it :-
I. Sequential – This is the method in which we search the data sequentially or line
by line until you find the desired data. E.g.. Magnetic tape, etc.
II. Direct – This is the method in which computer can go directly to the information
that the user wants. e.g. magnetic disk, optical disk, etc
 Common terminology related to memory

 A memory location always has a unique identifier called the address. Each memory unit is
a cell which stores 1 bit of information. Therefore a D-FF is also a 1-bit memory.
 A memory word is a collection of bits to form a multi-bit unit that carries information. A
word can be defined to be 8-bit, 16-bit, 32- bit or 64-bit. Other word sizes are possible, but
not common. An 8-bit word is known as a byte, which is the most common. For example,
when you were sending or receiving the ASCII character ‘#’ in Lab 1 or Lab 4, you were
using a byte with the value 0x23 (0x … is the way to indicate the number following the
prefix is a hexadecimal).
 Memory capacity (i.e. How much memory is in a memory chip?) is measured in kb, Mb,
Gb, Tb when measured on a chip. However, these memory bits could be organised in
different way. For example, a 8Mb chip could be organised as 1M of 8-bit words, 2M of 4-
bit words, or 8M of 1-bit storage. The number of unique words in a memory module
determines how many bits of address information is required to uniquely specify each
location uniquely.
 When using memory, we could be reading or fetching information from memory. This is a
READ operation.
 We could also be storing information to the memory. This is a WRITE operation.
 In any computer systems, storage of information is always organiised in a
hierarchy as shown here as a triangle. This has many layers representing
different type of storage.

 At the bottom is remote storage such as storage in the cloud (e.g. dropbox or
iCloud), or central file server offered by an organisation for its employees. This is
the largest and cheapest, and normally the slowest form of storage. This storage
could be very large.

 The next level is local disk storage, which is faster than the remote storage, but
slower than the next layer up. Most disks are now 128 GB or more. (We use GB
for giga byte and Gb for giga bit.)
 The next level is the main memory or DRAM in the computer. Modern
computer would come with 2GB or more of main memory. These normally
come on small PCBs and are swappable. In that way, one could “upgrade”
the memory, meaning that you can add more to the system.

 The next two levels are SRAMs on the processor chip itself. They are L2 (level
2) and L1 (level 1) cache memory. Cache memory is usually not large, but
has much faster access than all the other types of memory. Information are
generally fetched in “cache lines” which is multiple bytes.

 Finally the fastest memory are the registers inside the Central Processing Unit
(CPU).

 Information moves up and down this hierarchy in these order, going through
each layer.
MEMORY MAIN CHARACTERISTICS

 Access time – how fast can one perform a read or write operation
 Volatility – volatile memory are those that looses information when power is
removed.
 Non-volatile memory are those that keep information even when no power
is applied.
 Power consumption – the lower the better
 Density – how many bits can be stored?
 Cost – this is normally directly related to capacity and hence silicon area of
the chips.
Key Differences Between RAM and ROM

 RAM is a temporary storage type of memory as data lasts only till the time the
power supply is on. On the contrary, ROM is a permanent memory that retains the
data for a longer duration.
 Random-access memory is volatile in nature. But read-only memory is non-volatile
in nature.
 The operating speed of RAM is faster than the ROM.
 RAM is classified as the primary memory of the system. However, ROM is
categorised as the secondary memory.
 RAM offers memory capacity in GB usually 1 to 256 GB per chip. On the contrary,
ROM permits the storage capacity in MB, usually in the range 4 to 8 MB per chip.
 RAM stores data on transistors thus requires a continuous source of power. While
data stored in ROM remain unaffected with power failure.
 The data stored in RAM is changeable by the user. Whereas the data in ROM cannot
be altered by the user.
 RAM offers the user to read as well as write data but data in ROM is pre-written and
thus can only be read by the user.
 The processor can directly access the data present in RAM. However, the data present
in ROM do not permit direct access to the processor.
 The cost of Random access memory is comparatively higher than the read-only
memory.
 Programmable read-only memory (PROM), or one-time programmable
ROM (OTP), can be written to or programmed via a special device called a
PROM programmer. Typically, this device uses high voltages to permanently
destroy or create internal links (fuses or anti-fuses) within the chip.
Consequently, a PROM can only be programmed once.
Types of EPROMs

 There are two main types of EPROM technology:

1. UV-EPROM. A UV-EPROM is programmed at a high voltage. As the name
indicates, it is erased by shining UV light through the window on the chip. Most UV-
EPROM chips can store data for 20 or so years and can accommodate a high
number of read cycles.
2. One-time programmable (OTP) EPROM. An OTP EPROM is similar to a UV-
EPROM with one notable exception: It does not have a window for UV light and is,
therefore, not readily erasable. However, in some cases, OTP EPROMs can be
erased using X-rays.
 EPROM replaced ROM and PROM. After about 30 years of continued use,
EPROMs were eventually replaced by electrically erasable programmable
read-only memory (EEPROM) and flash memory.
 EEPROM is essentially an updated version of EPROM technology -- the key
difference is that EEPROM chips are electrically erasable. EEPROMs require
a high voltage for reprogramming, which can be generated on the chip by
a charge pump. On these chips, writes require much more power than
reads. Moreover, write times can be slow, and the memory itself supports a
limited number of write cycles. Nonetheless, EEPROMs are extensively used
in modern-day computers along with flash memory.
 One reason for their wide application is that an EEPROM and flash can be
erased and reprogrammed in place on the circuit board itself.

 In contrast, EPROM chips were programmed on an external programming

device and then placed on the computer's circuit board.

 Further, in an EEPROM, the data is erased using electrical signals, which is

much more convenient than using the UV light that's required to erase and
reprogram EPROMs.
 Dynamic random access memory (DRAM) is a type of random access
memory that stores each bit of data in a separate capacitor within an
integrated circuit. Since real capacitors leak charge, the information
eventually fades unless the capacitor charge is refreshed periodically.
Because of this refresh requirement, it is a dynamic memory as opposed to
SRAM and other static memory
Static Memory Cell

 A typical static memory cell

comprises a latch (2 cross-
coupled inverters) and 2 access
transistors
 The access transistors are turned
on when the word line is selected
 The complementary bit lines are
connected to the latch when the
cell is selected
Types of DRAM

 Asynchronous DRAM
 Fast Page Mode (FPM) DRAM or FPRAM
 Extended Data Out (EDO) DRAM
 Burst EDO (BEDO) DRAM
 Synchronous Dynamic RAM (SDRAM)
 Direct Rambus DRAM (DRDRAM)
 Double Data Rate (DDR) SDRAM
 Flash memory is also a popular choice for use in modern computers.

 Unlike both EPROMs and EEPROMs, which can only write or erase 1 byte at
a time, flash chips are designed for multibyte operations. This means they
can both write and erase entire blocks of data in one operation, speeding
up the erasure or reprogramming process and minimizing the damage to
transistors.

 Damage is a common problem with EPROMs and, to a lesser extent,

EEPROMs due to multiple erasure and reprogramming cycles.
Difference between ROM & RAM