PS/2 connectors
One connector is for plugging in the mouse, this is the
green PS/2 connector. The other is for plugging in the
keyboard, which is the purple connector. But PS/2
connectors are an older technology and is slowly being
phased out in favor of a USB port.
USB ports
The most common interface on a motherboard is the
USB port. USB stands for universal serial bus, and
motherboards would typically have several USB ports
because there are so many different peripherals that
utilize the USB interface, such as keyboards, mice,
cameras, external drives, and printers. In addition to
connectivity, the USB port also supplies electric power
to that specific peripheral. Some USB ports are
mounted on the rear input-output panel and some are
located directly on the surface of the motherboard.
The first USB interface was USB 1.0 in 1996. The
transfer speed was 1.5 Mbit/s. USB 1.1 was
introduced in 1998, with a transfer speed of 12
Mbit/s. USB 2.0 came out in 2001, with a transfer
47
speed of 480 Mbit/s. The latest version is USB 3.0,
with data transfer speeds of up to 5 Gbit/s.
Serial port
Another interface that can be found in older
motherboards is the serial port. The term serial refers
to sending data one bit at a time. The serial port is an
older technology interface which is rarely seen on new
motherboards today. Back then, this was mainly used
for connecting terminals and modems to computers, but
now it has been widely replaced by the faster USB
interface. The most common interface of a serial port is
the RS-232 standard, which uses the common "D"
connector such as the DB-9.
Parallel port
The parallel port is another interface that could be
found on a motherboard's I/O panel. It was mainly
used for connecting printers, and like the serial port, it is
also being phased out (pretty much gone) and replaced
by the faster performance of the USB port. It uses a
wide D-sub connector known as the DB-25. And unlike
serial ports, that send data one bit at a time, a parallel
48
port sends data signals simultaneously over several
parallel channels.
Integrated video adapter
A video adapter is another interface that could appear
on the motherboard. Now, I say it could appear,
because some motherboards have a video adapter and
some don't. But the motherboards that do have a video
adapter, then this is known as integrated video,
because the video adapter and the motherboard are
essentially one unit. The video adapter is what
generates images from your computer to your monitor.
The most common kind of port on an integrated video
adapter is VGA, which stands for video graphics
array. The VGA port carries analog data. It has 15
pins divided into 3 rows and usually has a blue color.
VGA video adapter
Integrated video adapters are usually not very powerful.
They are good for normal everyday use with light
applications, but when used for extensive graphic
applications such as gaming, they can fall short.
49
That's why a lot of people will bypass the integrated
video and add an expansion video card that has
enough power to suit their needs.
Expansion video card
Firewire port
The IEEE 1394 connector is commonly known as
firewire. Firewire is recognized by its "D" shape, and is
commonly used to attach devices such as digital
cameras and printers. It's very similar to a USB port.
Firewire and USB are both used to attach peripherals
and they both have similar transfer speeds, but firewire
is not nearly as popular as USB. That's why on a
modern motherboard today, there might not be any
firewire ports, or if there are, you might find one or two.
Firewire has a transfer speed or 400 Mbit/s.
The NIC or network interface card is a port on the
motherboard that is used for networking purposes, such
as connecting to the internet and sharing data between
50
NIC port
computers. The port is designed
for an Ethernet cable with an
attached RJ-45 connector. A
NIC provides a computer with a
constant dedicated connection to
a network. Every NIC has its
own unique identifier, called a
RJ-45
MAC address. The transfer
connector
speed ranges from 10 - 1000
Mbit/s.
Integrated sound port
A sound card is another type of interface that could be
found on a motherboard. And just like integrated
video, if I motherboard has a sound card built-in, then
this is known as integrated sound. A sound card is
what processes audio through the computer's
speakers. A basic sound card has an audio output
port for attaching speakers, and an import port for a
51
microphone for recording purposes. More
sophisticated sound cards will have extra ports. For
example, for subwoofers, surround sound, and ports for
other digital audio equipment.
eSATA port
Some motherboards have a SATA port on the rear I/O
panel. This is known as eSATA or external SATA.
It's used for attaching an external SATA device to a
computer, such as an external SATA hard drive. It
functions similar to USB and firewire, but the transfer
speed is faster. The disadvantage it has compared to
USB and firewire, is that eSATA requires a separate
power plug to supply the drive with power. However,
there is a new SATA port called eSATAp, which is
power over eSATA. This port combines data transfer
and power in an all-in-one port, similar to USB and
firewire.
52
Adapter/ Expansion Cards
Adapter cards or expansion cards are circuit boards
that can be installed into the expansion slots on a
computer's motherboard. These adapter cards are
installed to increase the functionality of the computer.
Some examples of expansion cards are video cards,
audio cards, and storage cards.
Motherboard with 2 adapter
cards installed.
One of the most common adapter cards is a video
card. A video card is what generates images from your
computer to your monitor. A typical video card is a
printed circuit board that directly attaches to the
computer's motherboard. Video cards are also known
as graphics cards and graphics adapters.
Video card
53
A video card has several key components, such as a
graphics processor unit or GPU, memory chips, a bus
type, and video ports. The way video is transferred
from the video card to the monitor is through the video
ports. GPU
Memory
Video ports
Bus type
There are several different types of video ports that are
used today. One port is called S-Video or separate
video or also known as super video. S-Video is an
analog transmitter. It transmits two signals over one
cable. One signal is for color and the other signal is for
brightness. The S-Video port is round and is usually a
black color.
S-Video
Another port is called VGA, which stands for video
graphics array. The VGA port is an older technology
that was developed in 1987. The VGA port has 15 pins
54
divided into three rows and usually has a blue color.
The VGA port carries analog data.
VGA port
Another video port is called a digital visual interface or
DVI. This is a newer technology designed to succeed
the older VGA port. It was developed in 1999 and it was
designed to provide uncompressed high-quality video to
LCD monitors.
DVI port
Now there are three different
versions of the DVI cable
standards. There's DVI-A, where
DVI-A
the A stands for analog. This is
used to send only analog signals.
There's DVI-D, where the D
DVI-D (single link) stands for digital. This is used to
send only digital signals.
DVI-D (dual link)
55
There's also the DVI-I, where
the I stands for integrated. This
DVI-I (single link) is used to send both analog and
digital signals.
DVI-I (dual link)
Now on the connectors that are able to send digital
signals, which would be the DVI-D and DVI-I, there are
two different options in the DVI standard. There is
single link and dual link. The difference is, that dual
link has six extra pins, and these pins are what allows
for a higher resolution than single link cables.
Another type of video port is called HDMI, which
stands for a high definition multimedia interface.
HDMI was developed in 2002 and it was designed for
transmitting uncompressed video and audio digital
data through a single cable. HDMI is one of the best
standards for high definition in consumer electronics,
delivering crystal clear video, as well as audio.
HDMI port
And there is also the DisplayPort. The DisplayPort
debuted in 2006 and it was developed by VESA, which
stands for the video electronics standard association.
56
The DisplayPort was primarily designed to be used for
video, but in addition to video, it can also be used to
carry USB and audio data as well. It's a high-
performance interface that is meant to replace the
older VGA and DVI interfaces. And it can also connect
using adapters to the older DVI, VGA, and HDMI ports.
DisplayPort
The sound card is another type of adapter card that
processes audio through the computer speakers. A
sound card attaches into the computer's motherboard
through a bus slot. A basic sound card has an audio
output port for attaching speakers and an input port
for a microphone for recording purposes. More
sophisticated sound cards have extra ports for
example ports for subwoofers and surround sound.
Sound card
57
Some people have video capture cards installed on
their computer. These cards allow a user to capture
analog video, such as from a video camera. Then it will
convert it to a digital form and then it can be stored on
their computer's hard drive.
Video capture card
TV tuner cards allow signals from a television to be
picked up by a computer. So you can not only watch TV
from your computer, but you can also record TV
programs and then store them digitally in your
computer. In fact, a lot of TV tuner cards also act as a
video capture card.
TV tuner card
60
A wireless network card does the same thing that a
wired card does, except that a wireless card does not
use a cable. It instead uses a wireless connection. It
has a built-in antenna that is used to wirelessly connect
to a network. Now, this is very convenient for computers
and devices that have a wireless network card because
they can be placed anywhere in different locations in a
home or office without the hassle of messing with
cables.
Wireless
network card
61
RAM (memory) Slots
RAM or random access memory is temporary
storage memory that's installed on the motherboard in
the memory slots.
Memory slots
The motherboard can have a various number of
memory slots. The average motherboard will have
between two and four slots. Memory slots come in
different types depending upon what type of RAM it
accepts.
RAM DIMM
For example, most motherboards accept DIMMs
because it's the most common type to date. DIMM
stands for dual inline memory module. A DIMM has
168, 184, 240, or 288 pins. A DIMM is a dual inline
module because it has two independent rows of these
62
pins - one row on each side. DIMMs will also have a
64-bit data path (which we will discuss shortly).
There is also the SIMM, which stands for single inline
memory module. SIMMs are an older technology and
are not produced anymore. SIMMs with either have 32
or 72 pins, and they have a 32 bit data path.
SIMM
The term 32 or 64 bit data path, refers to the number of
bits of data that are transferred in one cycle. The more
bits that are transferred in one cycle, the faster the
computer will be. A single bit or one bit of data is the
smallest form of data that the computer reads.
Because in the computing world, a computer only
understands ones and zeros, which is represented by
a single bit of data.
Now there's also the term byte, and 8 bits = 1 byte.
So if a memory DIMM is rated to have a 64 bit data
path, then that means that it has an 8 byte wide data
path or bus, because 64 / 8 = 8. If a memory SIMM
that is rated to have a 32-bit data path, then that
means that it has a 4 byte wide bus, because 32 / 8 =
4. So that's why DIMMS are faster than SIMMS. Now
this information will be useful in an upcoming lesson on
how we determine the exact bandwidth (speed) of
memory modules.
63
RAM
In a previous section, we talked about secondary
memory, which is permanent storage. This dealt with
hard drives. Now we're going to talk about primary
memory or temporary storage, and this is called RAM.
RAM stands for random access memory. In order for
data or a program to run, it needs to be loaded into
RAM first.
So here is how it works. The data or program is stored
on the hard drive. Then from the hard drive, it is
loaded into RAM.
Data from the hard drive is
being loaded in to RAM.
Once the data is loaded into RAM, the CPU can now
access the data or run the program.
64
Once the data is in RAM,
the CPU can now access it.
Now a lot of times if the RAM (memory) is too low, it
might not be able to hold all the data that the CPU
needs. So when this happens, then some of the data
has to be kept on the slower hard drive to compensate
for low memory.
From RAM to the
CPU
From the HD to
RAM
So instead of data going from RAM to the CPU, it has
to do extra work by going back to the hard drive, and
when this happens it slows down the computer. So to
65
solve this problem, all you need to do is increase the
amount of RAM on the computer. And by increasing
the memory, all the data can be loaded into RAM
without the need of constantly accessing the hard
drive. Therefore the result is a faster performing
computer.
RAM requires constant electrical power to store data
and if the power is turned off, then the data is erased.
RAM is stored on the motherboard in modules that are
called DIMMs, and these DIMMs come in different
memory sizes. Today they range anywhere from
128MB - 64GB of memory per DIMM.
RAM also comes in different types, such as dynamic
RAM or DRAM. DRAM is memory that contains
capacitors. A capacitor is like a small bucket that
stores electricity, and it's in
these capacitors that hold the
bit of information, such as a
one or a zero. Because that's
how computers read data,
which are ones or zeros. And
because DRAM has
capacitors, they have to be
refreshed with electricity
Capacitor
constantly, because capacitors
do not hold a charge for very
66
long, they constantly leak. And this refreshing of
electricity is where we get the name dynamic. The
capacitors have to be dynamically refreshed often
otherwise they will forget the information that they're
holding.
SRAM stands for static RAM. This memory uses
transistors instead of capacitors, so it does not have to
be refreshed unlike DRAM. Therefore it is much faster
than DRAM, but it's also very expensive. An example
SRAM would be the memory cache levels that are
used by the CPU.
Another type of memory is called SDRAM, which
stands for synchronous DRAM. This type of memory
is what is used today in RAM DIMMs. The difference
between SDRAM and DRAM, is basically speed. The
older DRAM technology operates asynchronously with
the system clock, which basically means that it runs
slower than the system clock, because its signals are
not coordinated with it. However SDRAM runs in sync
with the system clock, which is why it is faster than
DRAM. All the signals are tied to the system clock for
a better controlled timing.
DRAM - Dynamic RAM - Operates
asynchronously with the system clock.
SDRAM - Synchronous DRAM - Operates
synchronously with the system clock.
67
SDRAM is rated at different speeds. For example, a
stick of old SDRAM, way back in the 1990's, could be
labeled PC-100. The 100 equals the speed at which it
operates, which is 100 MHz. And since SDRAM only
comes in 64 bit modules, it has an 8 byte wide bus,
because 64 / 8 = 8. So to figure out the total
bandwidth (speed) of PC-100, you multiply 100 MHz x
8 bytes which equals 800 MB/s. So the total
bandwidth of PC-100 = 800 MB/s.
PC-100
100 MHz = The speed at which it operates.
8 byte wide bus.
100 MHz x 8 bytes = 800 MB/s
So an SDRAM module labeled PC-133, you multiply
133 x 8 which equals 1066. So the total bandwidth for
PC-133 = 1066 MB/s.
68
As technology increased and processor and bus
speeds have gotten faster. A new RAM technology
was developed to keep up with a faster speeds of
computers. This newer technology is called DDR,
which stands for double data rate; and that's basically
what DDR does. DDR sends double the amount of
data on each clock signal, when compared to non DDR
RAM. Non DDR, or single data rate RAM, uses only
the rising edge of the signal to transfer data. However
DDR uses both the rising and falling edges of the clock
signal to send data. Which makes DDR twice as fast.
non DDR
DDR
DDR uses both the rising and falling edges of the
clock signal to send data.
DDR is also labeled differently than non DDR RAM.
Instead of including the clock speed in its name, like PC-
133, where 133 equals the clock speed, DDR uses the
total bandwidth instead. For example a DDR DIMM
labeled PC-2700, the 2700 is not the clock speed, but it's
the actual total bandwidth. The clock speed for PC-2700
69
is 333 MHz. So 333 MHz x 8 bytes is rounded off to
2700 MB/s, which is where we get the name PC-2700.
DDR uses the total bandwidth in its name.
PC-2700
333 MHz = the speed at which it operates.
8 byte wide bus.
333 MHz x 8 bytes = 2700 MB/s
Another example is PC-3200. PC-3200 has a clock
speed of 400 MHz. So 400 MHz x 8 bytes = 3200 MB/s,
which is where we get the name PC-3200.
A new technology that has succeeded DDR is DDR2.
DDR2 is faster than DDR because it allows for higher
bus speeds, and it also uses less power than DDR. A
DDR2 DIMM has 240 pins, compared to 184 pins on
DDR. Some examples of DDR2 are PC2-3200 and
PC2-4200.
DDR2 DIMM
70
An even newer technology is called DDR3. DDR3 is
twice as fast as DDR2, with a bandwidth of over 12800
MB/s. Like DDR, a DDR3 DIMM also has 240 pins, but
the notches in the DIMMs are in different places. So
you can't put a DDR3 DIMM in a RAM slot made for
DDR2. Motherboards are made to support a certain
type of memory, so you can't mix DDR, DDR2, or DDR3
on the same motherboard. Some examples of DDR3
are PC3-8500 and PC3-12800.
DDR3 DIMM
To meet the higher demands of faster processors and
memory controllers, a new technology was developed
called dual channel mode.
Dual channel mode
requires a pair of
identical DIMMs
installed on the
motherboard,
which allows the Dual channel memory slots
memory controller
the ability to communicate with 2 DIMMs
simultaneously. Therefore increasing the speed of
accessing the memory. In order for dual channel
mode to work, the motherboard must be equipped to
71
work with dual channel mode. The memory DIMMs
must also be identical to each other in speed, size, and
features. Then the DIMMs must be inserted into the
motherboard in a specific slot configuration in order to
enable dual channel mode. Typically the memory slots
will be color-coded to help assist in identifying where
they should be inserted. For example, above we have
some dual channel memory slots. So in order for dual
channel mode to work, you need to install a pair of
identical DIMMs in the slots of the same color. In this
case we put a pair of DIMMs in the yellow slots.
2 DIMMs installed in the yellow slots. Dual channel mode
is now enabled.
There is also a triple channel mode. Triple channel
mode is not very common and very few motherboards
offer this feature. Triple channel mode allows the
memory controller the ability to communicate with 3
DIMMs at the same time. In modern computers,
motherboards have a 64-bit architecture. So in single
channel mode, it can transfer 64 bits of data at a time.
In dual channel mode, that is doubled to 128 bits at a
time. In triple channel mode, it's tripled to 192 bits at a
72
time. So as stated before, triple channel mode is only
available on a few motherboards and only certain Intel
Core i7 processors support triple channel mode. So
here is an example of a
triple channel capable
motherboard. On this
motherboard there are
six memory slots with
two different colors. So
if you're going to install
three DIMMs on this
motherboard, you must
put the DIMMs in the
same color-coded slots
in order to utilize triple
Triple channel equipped channel mode. Those
memory slots.
DIMMs must also be
identical to each other.
The term single sided or double sided RAM doesn't
necessarily refer to the physical location of the memory
chips that are on the sides of the memory module.
Top view of DIMMs showing memory chips.
It instead refers to the groups of memory chips that a
memory controller accesses. So for example, double
sided RAM has two groups of memory chips. Now, this
73
doesn't mean that the memory chips are physically
located on both sides of the memory module. Now,
they can be on both sides of the memory module, or
they can only be on just one side. But that's not what
double sided means. Double sided means that the
memory controller sees these two groups of memory
chips separately, and it can only access them one
group at a time.
Single sided RAM has one group of memory chips.
These chips can be physically located either on one
side or both sides of the memory module, but that's not
the point. The point is, that because it's single sided
RAM, the memory controller can access it as one
group. Therefore since it's one sided, or one group,
the memory controller can access it as one group,
which makes single sided RAM faster than double
sided RAM.
Some RAM modules have ECC, which stands for
error correcting code. ECC detects if the data was
correctly processed by the memory module, and
makes a correction if it needs to. You can tell if a
RAM module has ECC by counting the number of
memory chips on the module.
non ECC 8 chips
ECC 9 chips
74
In a standard non ECC DIMM, it will have eight
memory chips, but in an ECC module, it will have nine
memory chips. Most RAM modules today are non
ECC, and this is because of the advancing technology
that has minimize memory errors and has made non
ECC RAM more stable. Typically today, ECC memory
is mostly used in servers, because servers need to be
up and running at all times, and using ECC memory is
just an extra precaution to guard against any memory
errors.
There is also buffered and unbuffered RAM. Now
buffered RAM, which is also called registered RAM, is
made to add stability to RAM. Buffered RAM adds an
extra register between the RAM and the memory
controller. The extra register stores data, or buffers the
data, before it gets sent to the CPU. This is what adds
stability and reliability in computer systems that have a
lot of memory modules installed. So it's basically used
to lessen the electrical load on the memory controller,
that’s produced when a computer uses a lot of memory
modules, for example in servers.
The kind of RAM that is used in smaller devices, such
as laptops, is called SODIMM. SODIMM stands for
small outline dual inline memory module.
SODIMM
75
SODIMMs are roughly half the size of regular DIMMs,
and like DIMMs, that are used in desktops, SODIMMs
also come in different types, such as DDR and DDR2
and DDR3. So if you plan on adding RAM to your
laptop, you need to make sure you install the correct
kind, because the different levels of DDR in SODIMMs
are not compatible with each other.
76
Cooling
Cooling is very important to a
computer. Computers
generate a lot of heat and if
the components of the
computer are not adequately
cooled, the computer will
overheat. If the computer
overheats, the computer could
run slow, lock up, or shut
down. It will also eventually shorten the life of the
computer because heat is an enemy to a computer's
longevity. Two of the biggest heat generators come
from the CPU and the video card.
So the main way to adequately cool the computer is
with case fans.
Exhaust
fan
Intake
fan
Case fans inside a tower computer case.
77
Case fans are mounted inside the computer case. A
basic case fan setup will include at least two fans. The
fan that is mounted in the front of the computer case,
is the intake fan. The intake fan is for drawing cool air
from outside the case, to inside the case. The other
fan is the exhaust fan, which is located in the back of
the case. The exhaust fan is designed to push warm
air outside the case. Once these fans are in place,
they are designed to work with each other to create a
constant flow of cool air coming into the computer, to
cool the computer components.
Warm air
Cool air
Case fans creating a constant flow of cool air.
So the cool air comes in and cools the computer
components. Then as the air makes contact with the
hot components, the air naturally gets warmer, and is
then drawn outside the case. Then the cycle starts all
over again to create a constant circulation. This type of
cooling is known as active cooling.
78
The biggest heat
generator in the computer
is the CPU. The CPU is
the brain of the computer,
and it produces an
enormous amount of heat
in a very short amount of
time. In fact, if the CPU
were to run by itself
The CPU is the biggest
heat generator. without any extra cooling
components, the CPU
would likely fry itself within 10 seconds (trust me, I
tried it). So to remedy this problem, a CPU needs a
heat sink to help dissipate the heat. A heat sink is
basically an aluminum block with fins that directly
makes contact with the CPU.
The heat sink's purpose is
to increase the surface
area of the CPU so that it
can make more air contact
for cooling. The larger the
heat sink, the larger the
surface area will be,
therefore increasing the Heat sink
cooling ability.
The fins on the heat sink are designed to further
increase the surface area for air circulation. Once the
heat sink makes contact with the CPU, the heat will
transfer from the CPU to the heat sink, where the air
can cool the heat sink, which will cool the CPU. This
type of cooling is known as passive cooling.
79
The heat sink fins are designed to further increase the
surface area for air circulation.
It's important that the heat sink and the
CPU make the most contact with each
other so that adequate cooling can
take place. So that's why it's important
to apply thermal compound on the
CPU before attaching the heat sink.
Thermal
compound
being applied
on a CPU.
Thermal compound is used to fill in the microscopic air
gaps between the heat sink and the CPU, to make up
for the imperfections of the flat surfaces. The surface
areas between the CPU and heat sink are flat, but
they are not perfectly flat when examined with a
microscope.
80
Magnified
Microscope reveals air gaps. The surfaces
are not perfectly flat.
Thermal compound is designed to fill in the
microscopic air gaps so the most contact can be made
between the heat sink and CPU.
Magnified
After thermal paste is applied. It fills in the
microscopic air gaps.
Another type of cooling is water cooling. Instead of
using air, this type of cooling uses water to cool the
computer components. For example, here is a water
cooling unit for the CPU. There is a pump, hosing,
and a radiator.
81
Hosing
Radiator
Pump
CPU water cooler
Inside this unit is water. The unit is placed directly on
top of the CPU just like a traditional heat sink and the
pump inside constantly circulates the water throughout
the entire unit to keep the CPU cool. Once the water
reaches the radiator, the water is air cooled by a
radiator mounted fan that draws air into the radiator
and cools the water. Then the cycle is repeated, so a
constant flow of cool water makes contact with the
CPU. So as a result, water cooling units, cool
components far better than air cooling. They are also
much quieter, but at the same time, water coolers are
more expensive.
82
CPU, Socket, &
Chipset
The CPU socket is the place on the motherboard
where the CPU is placed. The socket is a square
plastic or metal holder with multiple holes to
accommodate the pins on the bottom of the CPU.
CPU
Socket
As these holes and pins make contact, they provide
physical and electrical contact between the
motherboard and the CPU.
A modern CPU socket is called a ZIF, which stands
for zero insertion force, which basically means that
the CPU is installed in the socket with no force.
The CPU just drops in the
socket easily. There are
different types of CPU socket
designs called packages.
One of the most common
types of these packages is ZIF socket
the PGA or pin grid array.
83
The PGA package is a typical square design with
holes and a lock down lever.
The latest in socket design
packages is called LGA, which
stands for land grid array.
The LGA socket is a metal
casing with a door that closes
over the CPU, and locks down
with a lever. Unlike previous
socket versions, which have
holes, the LGA has pins that
make contact with the bottom
of the processor. LGA
LGA socket
processors don't have pins,
instead they have pads that
rest on the LGA socket pins.
For the CompTIA A+ exam, you're going to need to
know certain characteristics of several socket types.
These socket types are categorized by two different
brands. These brands are Intel and AMD.
So starting with the Intel sockets,
which use the LGA package.
The first one that we're going to
talk about is the LGA 775 socket.
This is also known as socket T. It
was released in 2004, and it has
775 pins as its name states. The
LGA 775 socket
LGA 775 was the successor to
socket 478 and it was designed for the Pentium 4
and Pentium dual core processors.
84
Next is the LGA 1366. This, as
its name states, has 1366 pins
and is also known as socket B.
It was released in 2008 and has
succeeded the LGA 775. The
LGA 1366 uses the Intel Core i7
LGA 1366 socket and Xeon processors.
The LGA 1156 is also known as
socket H or socket H1. This was
released in 2009 and has 1156
pins. This was the first socket to
be used by the Intel core i3 and i5
processors.
LGA 1156 socket
The LGA 1155 is also known as
socket H2. This was designed to
replace the LGA 1156 and has
1155 pins, which is 1 less than
the LGA 1156. But the CPUs
designed for the LGA 1155 and
LGA 1156 are not compatible
LGA 1155 socket because the notches in the
sockets are different. It was
released in 2011, and these were designed for Intel
CPUs that use the Sandy Bridge and Ivy Bridge
architecture.
85
The LGA 1150 is also known as
socket H3. This has 1150 pins
and was released in 2013. The
LGA 1150 supports Haswell and
Broadwell based microprocessors
and has succeeded the LGA
1155. LGA 1150 socket
The last of the Intel sockets is the
LGA 2011, which is also known
as socket R. The LGA 2011 has
2011 pins and was released in
2011. It has succeeded the LGA
1366 and was designed for high
performance CPUs that are
LGA 2011 socket based on Sandy Bridge and Ivy
Bridge processors.
The next group of sockets is the
AMD sockets, which utilize the PGA
package. So the first socket is the
AM3. Socket AM3 was released in
2009 and is the successor to the
AM2+. The AM3 has 941 pins.
AM3 socket
Next is socket AM3+, which is the
successor and a modification of the
AM3. It has 942 pins and was
released in 2011. The AM3+ does
retain some compatibility with AM3.
AM3+ socket
86
So CPUs designed for AM3 will work in AM3+ sockets.
The next socket is FM1. This
was released in 2011 and has
905 pins. These were designed
for AMD APU processors. APU
stands for accelerated processing
unit. These were designed to act
as a CPU and a graphics
FM1 socket
accelerator, on a single chip.
Next is FM2. This has 904 pins and was released in
2012. There is also FM2+. This was released in
2014. FM2+ has 906 pins and was a new revision to
the FM2.
FM2 socket FM2+ socket
The central processing unit or CPU
is the main component on the
motherboard. It's the brain of the
computer where all the data
processing takes place. It's in
charge of executing program
instructions and logical calculations.
87
The CPU is the largest component on the
motherboard. It's a square chip that is inserted into
the motherboard, in a plastic or metal holder called a
CPU socket. Directly on top of the CPU is the heat
sink and fan, and these are used to keep the CPU
from overheating.
Fan
Heat sink
CPU
CPU socket
The speed of the CPU is measured in megahertz
(MHz). For example, 1 MHz equals 1 million cycles
per second. 500 MHz equals 500 million cycles per
second. 1 gigahertz (GHz) equals 1 billion cycles per
second. Today's high-end processors average a
speed of over 3 GHz per second.
Inside the processor is the
core. The core is where the
reading and execution of
instructions take place. A
processor that has a single
core, processes instructions
one at a time. However, today's
higher-end processors will Single core processor
88
have multiple cores. These are called multi-core
processors, and they can process more instructions
than a single core processor. Which gives a multi-core
the ability to multitask and have a greater overall
performance. Some examples of multi-core processors
are, dual core processors which has two cores. Another
example is a quad core processor, which has four cores.
Dual core processor Quad core processor
Two of the biggest manufacturers of processors are
Intel and AMD. Intel is the largest manufacturer of
processors and was founded in the late 1960s and has
since dominated the CPU market for a number of years,
until the rise of AMD started to become its chief
competitor. Some of the Intel processors are known as
the 286, 386, 486, Celeron, Pentium, and Xeon
processors.
Intel CPU
89
Advanced Micro Devices or AMD, is the second largest
manufacturer of processors, and it was also founded in
the late 1960s. However, AMD didn't really start to
compete with Intel in the CPU market until the mid-
1990s. Some of the AMD processors are known as the
K5, K6, Athlon, Duran, Sempron, Athlon 64, Opteron,
Phenom, FX, and Ryzen.
AMD CPU
CPUs can come in 32 or 64-bit versions. The difference
between a 32-bit and 64-bit is the way that it handles
memory. The bit size of the CPU refers to the memory it
can address. A 32-bit CPU can reference 2³² bytes of
memory, which equals about 4 GB (gigabytes).
64
However, a 64-bit CPU can reference 2 bytes of
memory, which equals to about 16 exabytes, which is 4
billion times more memory than a 32-bit.
90
Now that number is so huge that it's a virtually unlimited
because we will never need to use that amount of
memory.
So going back to what we stated before. In order for
data or a program to run, it needs to be loaded into RAM
first. So the data is stored on the hard drive, and then
from the hard drive, it's loaded into RAM. Then once it's
loaded into RAM, the CPU you can now access the data
or run the program. Now in a 32-bit system, since a
maximum amount of memory it can support is 4 GB, it
may not be enough to hold all the data that the CPU
needs to make the computer run as fast as possible. So
when this happens, then some of the data has to be
kept on the hard drive to compensate for the low
memory.
CPU
CPU
From the HD to From RAM to the From RAM to the
RAM
32 Bit 64 Bit
So instead of data going from RAM to the CPU, it
has to do extra work by going back to the slower
hard drive. When this happens, it slows down the
computer.
91
However, on a 64-bit system, it's able to store a lot more
memory than 4 GB. Which means that more data can
be loaded into the faster RAM, than on the slower hard
drive; and because it can store more data on the faster
RAM than on the slower hard drive, the computer is able
to run a lot faster. So in a nutshell, this is why a 64-bit
system is faster than a 32-bit system.
There's also what's called memory cache (CPU cache).
The memory cache uses SRAM or static RAM, which is
very fast memory when compared to regular DRAM that
is used for primary memory. The memory cache is the
CPU's internal memory and its job is to hold data and
instructions waiting to be used by the CPU.
The memory cache rapidly assists in feeding the CPU
data, because RAM is still not fast enough for the CPU.
So basically what cache does, is that it holds common
data that it thinks the CPU is going to access, over and
over again. When the CPU needs to access certain
data, it always checks the faster memory cache first to
see if the data it needs is there, and if it's not, then the
CPU will have to go back to the slower primary memory
or RAM, to find the data it needs. So that's why memory
cache is so important, because if the CPU can access
92
what it needs on the faster memory cache, then the
faster the computer will perform.
The memory cache comes in different levels. For
example, there's level 1 cache, which is also called
primary cache. Level 1 cache is located on the CPU
itself. So it runs at the same speed as the processor.
So it's very fast and is the fastest memory cache on the
computer. There was also level 2 cache, which is also
called external cache. Level 2 cache is used to catch
recent data accesses from the processor that were not
caught by the level 1 cache. So in a nutshell, if the CPU
can't find the data it needs on the level 1 cache, it then
searches the level 2 cache for the data. Then if level 2
doesn't have it, then the CPU has to go to the next level,
which is level 3 cache. Level 3 cache is used to catch
recent data accesses from the processor that were not
caught by the level 2 cache. Then if level 3 doesn't
have it, then the CPU has to go back to RAM to find the
data it needs.
Level 3 cache is
located on the
processor.
Shared between all
In modern CPUs, the cores in the
level 2 cache is CPU.
located on the
processor.
Level 1, 2, & 3 cache inside the CPU.
93
Level 2 cache is generally located on a separate chip on
the motherboard, or in modern CPUs, it would also be
located on the processor. Level 2 cache is larger than
level 1 cache, but it's not as fast as level 1 cache. Level
3 cache is also located on the processor. Level 3 is
larger than level 2, but it's not as fast as level 2 cache.
Level 3 is often referred to as shared cache, because its
memory is shared between all the cores on the CPU,
whereas level 1 and level 2 cache are dedicated to their
own CPU core.
One of the main and one of the most important
components you'll find on the motherboard is the
chipset. Older motherboards were designed with a lot
of different chips, scattered all over the motherboard.
There were chips for
different things, like chips for
bus controllers, memory
controllers, keyboard
controllers, and so on. So
they had a lot of different
chips controlling different
functions on the
motherboard. So as
technology progressed,
Older motherboard with
computer engineers decided
chips scattered all over.
to reduce the number of
chips and have them more
in a centralized location. So instead of having these
different chips scattered all over the motherboard,
controlling different functions; they reduced the number
94
of chips to do the same job and
condensed them to only a few
chips, or what's now called a
chipset. And that's what a
chipset is, a chipset is a smaller
set of chips that has replaced a
larger amount of chips. The
chipset's job is to control data
flow between the CPU, the
peripherals, bus lots, and
memory. So all of the different
Modern motherboard
with a chipset. parts of the motherboard,
communicate with the CPU
through the chipset.
The chipset basically consists of two chips, one is
called the northbridge and the other is called the
southbridge. The northbridge is located in the upper
or northern part of the
motherboard, providing
you're looking at the
motherboard in the upright
position. It's located near
the CPU and is directly
connected to the CPU. It's
also directly connected to
the memory, and the
AGP(outdated) and PCI
express slots. So in order
for the CPU to communicate with the memory, and the
AGP or PCI express bus, it has to go through the north-
95
bridge first. So the northbridge acts like a
communication middleman between a CPU, AGP, or
PCI express, and memory.
The other chip is called the southbridge. The south-
bridge is located at the bottom or southern portion of
the motherboard, near the PCI bus slots. The south-
bridge connects to the PCI bus slots, SATA and IDE
connectors, and USB ports. So the southbridge is
responsible for the lower portion of the motherboard,
while the northbridge is responsible for the upper
portion. There is no direct connection between the
CPU and the lower portion of the motherboard. So if
the PCI slots, USB, IDE, or SATA ports needed to
communicate with the CPU, the information has to go
through the southbridge, then up through the north-
bridge, and then to the CPU.
Northbridge is Southbridge is
responsible for the responsible for the
upper portion of the lower portion of the
motherboard. motherboard.
96
The northbridge is faster than the southbridge. This is
because the CPU, PCI-E, and memory are the most
used and most important components of the
motherboard. So they need to operate at the highest
speed possible. The slower southbridge communicates
with a PCI bus, SATA and IDE connectors, and USB
ports, and these don't need to be as fast as the other
components. So basically the higher speed
components are connected to the northbridge and the
slower components are connected to the southbridge.
Now both the north and southbridge make these
connections to various parts of the motherboard using
pathways called a bus. A bus as simply a set of
pathways that allows data and signals to travel
between the components
on the motherboard.
The motherboard
contains several kinds of
buses that vary in speed
and bandwidth. So for
example, if a bus speed
is said to operate at 66
MHz, then that means
that particular bus can
send data at 66 million
cycles per second. The
higher the bus speed, Motherboard buses.
the faster the computer
can send data, which
97
Hyper-threading
improves the performance of the computer. A
motherboard's bus speed generally refers to the speed
of the front side bus. The front side bus is the
connection between the CPU and the northbridge
chipset.
Hyper-threading is a technology developed by Intel
that increases the performance of the CPU cores. It
enables multiple threads, which are sequences of
Bus Slots
instructions, to be run by each core to make the CPU
run more efficiently. And by doing this, the CPU can
perform more tasks in the same amount of time. So in
a nutshell, you can run a lot of applications at the same
time while maintaining the performance of your
computer when you have a hyper threaded CPU. In
other words your computer is not going to slow down.
All motherboards are equipped with input/output bus
slots. These are typically located on the bottom rear of
the motherboard. These bus slots are also called
expansion slots because these slots are used to
expand the capability of the computer.
98
The words you are searching are inside this book. To get more targeted content, please make full-text search by clicking here.
Discover the best professional documents and content resources in AnyFlip Document Base.
Search