EMMA HS3 Advanced Hardware Outline Week #4

useumpireSoftware and s/w Development

Dec 2, 2013 (3 years and 11 months ago)

129 views

EMMA HS3 Advanced Hardware Outline



Week #4



How Graphics Cards Work



Basics




Four Main Components





Motherboard
Connection


Interface to CPU

and Power





Graphics Processing Unit


Decide what to do with each Pixel on screen





Memory


Holds i
nfo about each pixel






Frame Buffer


holds completed images






Dual Ported


reads and writes at the same time






Memory Type






BIOS Chip





Monitor Connection






VGA (CRT/LCD)






DVI (Digital Visual Interface)




Performance




Frame rate


Frames per Second (FPS) Number of Complete images per second




Triangles or Vertices per second


Speed of Wire Frame Image




Pixel Fill Rate


Number of Pixels processed in a second




Hardware Specifications




GPU Clock speed (MHz)




Size of memory

bus (bits)




Amount of available memory (MB)




Memory clock rate (MHz)




Memory bandwidth ( GB/s)




RAMDAC speed (MHz)



New Egg
Graphics Card Specifications



Chipset Manufacturer


ATI & nVidia



Interface


PCI, AGP, PCI Express x1, PCI Express x1
6
, PCI

Express 2.1

x16



GPU
-

FireGL, Radeon, GeForce, FireMV

Memory Size


Up to 1
.792

GB



PixelPipelines vs

Stream Processors



Memory Interface


Up to 512 bit

bus size on graphics board



Memory Type


DDR, GDDR2, GDDR3, GDDR4
, GDDR5



Direct X req
uirements



I/O Connections


D
-
Sub, DVI, TV
-
Out, HDMI, ViVo



Max Resolution


up to 3840 x 2400



Crossfire Support


ATI paired graphics card support



SLI Support


Scalable Link Interface


nVidia


link two or more cards to single output



Cooler


Fan or Fanless
, water cooled



HDCP

-

high
-
bandwidth digital
-
content protection





Documentation
-

How Graphics Cards Work

Graphics Card Specification Definitions, Graphics Card GDDR Memory


Homework
-


Graphics Card

Quiz




Wish List Graphics Card





Gr
aphics Card GDDR Memory

RAM memory is also used on video cards to make the video memory circuit. Until recently the video
memory used the exact same technology as the system RAM memory that is installed on the
motherboard. High
-
end video cards, however, ne
eded memory chips faster than the ones used on
the
PC
. So the manufacturers decided to go for DDR2 and DDR3 technologies.

DDR2 and DDR3 memories used on video card have different charac
teristics than the DDR2 and
DDR3 memories used on the PC


especially the voltage. That’s the reason they are called GDDR2
and GDDR3 (the “G” comes from “Graphics”).

In our
DDR2 Memory T
utorial

we explained the differences between DDR and DDR2 memories. As
we mentioned there, one of the main differences is the voltage: while DDR works at 2.5 V, DDR2
works at 1.8 V. This leads to a lower power consumption and less heat.

GDDR2 memories con
tinue to work at 2.5 V. Since they run at higher clock rates compared to DDR
memories, they generate more heat. This is the reason why only a few video cards used GDDR2
memories


only GeForce FX 5700 Ultra and GeForce FX 5800 Ultra used this kind of
memory
.
Shortly after GeForce FX 5700 Ultra was released many video card manufacturers released a
GeForce FX 5700 Ultra using GDDR3 memories, maybe to lower the heat and power consumption
ef
fects.

GDDR3 memories can work at 2.0 V (Samsung chips) or at 1.8 V (chips from other manufacturers),
solving the heat problem. This is the reason why this kind of memory is used by high
-
end video cards.

DDR3 memories

are not released to
PCs

yet, but they will probably work at 1.5 V, being different from
GDDR3 memories.

GDDR
-
3 graphics memory is the next evolution of high
-
speed DDR SDRAM technologies that have
played a major role in enab
ling GPUs to drive complex geometries and character animations and
deliver visual effects on par with the hottest motion pictures. GDDR
-
3 graphics memory enables
higher memory clock frequencies at a lower power level with fewer components and less constrai
nts
on system designers.

The main advantage DDR2 has over good ole DDR memory is that it runs on a lower voltage, which
lowers the power requirements, and allows it to scale higher with a small latency penalty.

GDDR3 (Graphics Double Data Rate3) takes thi
s one step further, requiring less voltage than DDR2,
and scaling even futher (though with some latency penalty). While the motherboard industry is
making the transition from DDR to DDR2 memory, right now GDDR3 is only used on
graphics cards
.









How Graphics Cards Work

by
Tracy V. Wilson

and
Jeff Tyson

The images you see on your
monitor

are made of tiny dots called pixels. At most common resolution settings, a
screen displays over a million pixels, and the compu
ter has to decide what to do with every one in order to
create an image. To do this, it needs a translator
--

something to take
binary data

from the
CPU

and turn it into a
picture you can see. Unless a computer has graphics capability built into the
motherboard
, that translation takes
place on the
graphics card
.

A
graphics card's job is complex, but its principles and components are easy to understand. In this article, we
will look at the basic parts of a video card and what they do. We'll also examine the factors that work together
to make a fast, efficient graphic
s card.


The graphics card creates a wire frame image, then fills it in
and adds textures and shading.

Graphics Card Basics

Think of a computer as a company
with its own art department. When people in the company want a piece of
artwork, they send a request to the art department. The art department decides how to create the image and then
puts it on paper. The end result is that someone's idea becomes an actua
l, viewable picture.


Photo courtesy of
HowStuffWorks Shopper

The four main components of a graphics card are

connections
for the motherboard and monitor, a processor, and memory.

A graphics card works along the same principles. The
CPU
, working in conjunction with software applications,
sends
information about the image to the graphics card. The graphics card decides how to use the pixels on the
screen to create the image. It then sends that information to the
monitor

through a cable
.

The Evolution of Graphics Cards

Graphics cards have come a long way since IBM introduced the
first one in 1981. Called a
Monochrome Display Adapter

(MDA), the card provided text
-
only displays of green or white
text on a black screen. Now, the minimum st
andard for new
video cards is
Video Graphics Array

(VGA), which allows 256
colors. With high
-
performance standards like
Quantum
Extended Graphics Array

(QXGA), video cards can display
millions of colors at resolutions of up to 2040 x 1536 pixels.

Integrat
ed Graphics

Many
motherboards

have integrated graphics capabilities
and function without a separate graphics card. These
motherboards handle 2
-
D images easily, so they are ideal
for producti
vity and Internet applications. Plugging a
separate graphics card into one of these motherboards
overrides the onboard graphics functions.

Creating an image out of
binary data

is a demanding pro
cess. To make a
3
-
D

image, the graphics card first
creates a wire frame out of straight lines. Then, it
rasterizes

the image (fills in the remaining pixels). It also
adds
lighting
, texture and color. For fast
-
paced games, the computer has to go through this process about sixty
times per second. Without a graphics card to perform the necessary calculations, the workload would be too
muc
h for the computer to handle. The graphics card accomplishes this task using four main components:



A
motherboard

connection for data and power



A
processor

to decide what to do with each pixel on the screen



Memory

to hold information about each pixel and to temporarily store completed pictures



A
monitor

connection so you can see the final result

Processor and Memory

Like a
motherboard
, a graphics card is a printed circuit board

that houses a
processor

and
RAM
. It also has an
input/output system
(B
IOS)

chip, which stores the card's settings and performs diagnostics on the
memory
,
input and output at startup. A graphics card's processor, called a
graphics processing unit

(GPU), is

similar to
a computer's CPU. A GPU, however, is designed specifically for performing the complex mathematical and
geometric calculations that are necessary for graphics rendering. Some of the fastest GPUs have more
transistors than the average CPU. A GPU
produces a lot of heat, so it is usually located under a heat sink or a
fan. In addition to its processing power, a GPU uses special programming to help it analyze and use data.
ATI

and
nVidia

produce the vast majority of GPUs on the market, and both compa
nies have developed their own
enhancements for GPU performance. To improve image quality, the processors use:



Full scene anti aliasing

(FSAA), which smoothes the edges of 3
-
D objects



Anisotropic filtering

(AF), which makes images look crisper

Each compa
ny has also developed specific techniques to help the GPU apply colors, shading, textures and
patterns.

As the GPU creates images, it needs somewhere to hold information and completed pictures. It uses the card's
RAM

for this purpose, storing data about each pixel, its color and its location on the screen. Part of the RAM
can also act as a
frame buffer
, meaning that it holds completed images until it is time to display them.
Typically, video RAM ope
rates at very high speeds and is
dual ported
, meaning that the system can read from
it and write to it at the same time.

The RAM connects directly to the
digital
-
to
-
analog converter
, called the DAC. This converter, also called the
RAMDAC, translates the i
mage into an analog signal that the monitor can use. Some cards have multiple
RAMDACs, which can improve performance and support more than one monitor. You can learn more about
this process in
How Analog and Digital Recording Works
.

The RAMDAC sends the final picture to the monitor through a cable. We'll look at this connection and other
interfaces in the next section.

Input and Output

Graphics cards connect to the computer through the motherboard. The
motherboard

supplies power to the card and lets it communicate with the
CPU
.
Newer graphics cards often require more power than the motherboard can
provide, so they also have a di
rect connection to the computer's power supply.

Connections to the motherboard are usually through one of three interfaces:



Peripheral component interconnect

(PCI)



Advanced graphics port

(AGP)



PCI Express

(PCIe)

PCI Express is the newest of the three and provides the fastest transfer rates between the graphics card and the
motherboard. PCIe also supports the use of two graphics cards in the same computer.

Most graphics cards have two monitor connections. Often, one is a DVI connector, which supports
LCD

screens, a
nd the other is a VGA connector, which supports
CRT

screens. Some graphics cards have two DVI
connectors instead. But that doesn't rule out using a CRT screen; CRT screens can connect to DVI ports

through
an adapter.

Most people use only one of their two monitor connections. People who need to use two monitors can purchase
a graphics card with
dual head capability
, which splits the display between the two screens. A computer with
two dual head, PC
Ie
-
enabled video cards could theoretically support four monitors.


Photo courtesy of
HowStuffWorks Shopper

Th
is Radeon X800XL graphics card has DVI, VGA
and ViVo connections.

DirectX and Open GL

DirectX and Open GL are
application programming
interfaces
, or APIs. An API helps hardware and software
communicate more efficiently by providing instructions for
complex

tasks, like 3
-
D rendering. Developers optimize
graphics
-
intensive games for specific APIs. This is why the
newest games often require updated versions of DirectX or
Open GL to work correctly.

APIs are different from drivers, which are programs that allow

hardware to communicate with a computer's
operating system
.
But as with updated APIs, updated device drivers can help
programs run correctly.

In addition to connections for the mothe
rboard and monitor, some graphics cards have connections for:



TV display
: TV
-
out or S
-
video



Analog video cameras
: ViVo or video

in/video out



Digital cameras
:
FireWire

or
USB


ADC Connectors

At one time, Apple made
monitors t
hat used the
proprietary Apple Display
Connector (ADC). Although
these monitors are still in use,
new Apple monitors use a DVI
connection.

Some cards
also incorporate TV tuners.

Choosing a Good Graphics Card

A top
-
of
-
the
-
line graphics card is easy to spot. It has lots of
memory

and a fast
processor
. Often, it's also more
visually appealing than anything else that's intended to go inside a computer's case. Lots of high
-
performance
video cards are illustrated or have decorative fans or heat sinks.

But a high
-
end card
provides more power than most people really need. People who use their computers
primarily for
e
-
mail
, word processing or Web surfing can find all the necessary graphics support on a
motherboard

with integrated graphics. A mid
-
range card is sufficient for most casual gamers. People who need
the power of a high
-
end card include gaming enthusiasts and people who do lots of 3
-
D graphic
work.


Photo courtesy of
HowStuffWorks Shopper

Some cards, like the ATI All
-
in
-
Wonder, include
connections fo
r televisions and video as well as
a TV tuner.

Overclocking

Some people choose to improve their graphics card's
performance by manually setting their clock speed to a higher
rate, known as
overclocking
s. People usually overclock their
memory
, since overclocking the GPU can lead to overheating.
While overclocking can lead to better performance, it also voids
the manufacturer's warranty.

A good overall measurement of a card's performance i
s its
frame rate
, measured in frames per second (FPS).
The frame rate describes how many complete images the card can display per second. The human eye can
process about 25 frames every second, but fast
-
action games require a frame rate of at least 60 FPS
to provide
smooth animation and scrolling. Components of the frame rate are:



Triangles or vertices per second
:
3
-
D

images are made of triangles, or polygons. This measurement
describes how q
uickly the GPU can calculate the whole polygon or the vertices that define it. In general,
it describes how quickly the card builds a wire frame image.



Pixel fill rate
: This measurement describes how many pixels the GPU can process in a second, which
tran
slates to how quickly it can rasterize the image.

The graphics card's hardware directly affects its speed. These are the hardware specifications that most affect
the card's speed and the units in which they are measured:



GPU clock speed (MHz)



Size of th
e memory bus (bits)



Amount of available memory (MB)



Memory clock rate (MHz)



Memory bandwidth (GB/s)



RAMDAC speed (MHz)

The computer's
CPU

and
motherboard

also play a part, since a very fast graphics card can't compensate for a
motherboard's inability to deliver data quickly. Similarly, the card's connection to the motherboard and the
speed at which it can get instructio
ns from the CPU affect its performance.

Graphics Card Specification Definitions

The
pixel pipeline

was a component within
3D accelerators
, most prominently

prior to
DirectX

9. The term encompasses
one of a number of parallel processing pipelines within a
graphics processing unit

(GPU). Each pipeline processes
pixel
,
texture
, and frequently
geometric

data. Various GPUs had differing numbers of pixel pipelines, and larger numbers of
these pipelines increased the pixel/texel per clock performance of the accelerator
. This performance was measured in pixel
and texture fill
-
rate. Real
-
time 3D rendering performance scales well with additional parallelism because of the nature of
3
D graphics

functions.

Every image you see on a screen is made up of thousands of pixels. A pixel is a single point within an image, and is
normally capable of displaying either three colours (red, green, blue) or four colours (cyan, yellow, magenta, black
).
Pixels are associated with the screen resolution of your display, so if you were to play a game at a common resolution
such as 1280x1024, your display would show 1280 pixels across the screen, and 1024 pixels from top to bottom.

The pixel pipeline proce
sses the pixel, texture and geometric data received from the Vertex Shaders. Different GPUs
(Graphics Processing Unit) have different numbers of pixel pipelines, but as a rule of thumb, the more pipelines a graphics
card has, the faster the card can proces
s the data for rendering the images on
-
screen.


Each pixel is made up of a series of fragments, which are processed by the pixel shader according to calculations made by
the vertex shader. Once each fragment is processed it is held in a buffer where it is

built into a complete pixel by the
Raster Operator unit.

Pixel shading is usually the most intensive part of the graphics rendering process on a modern GPU and so usually takes
the most time.

Microsoft DirectX

is a collection of
application programming interfaces

for handling tasks related to
multimedia
,
especially
game programming

and video, on
Microsoft

platforms. Originally, the names of these APIs all began with
Di
rect, such as Direct3D, DirectDraw, DirectMusic, DirectPlay, DirectSound, and so forth. DirectX, then, was the generic
term for all of these Direct
-
something APIs, and that term became the name of the collection.


The
High
-
Definition Multimedia Interface

(
HDMI
) is a licensable audio/video connector interface for transmitting
uncompressed, encrypted digital streams. HDMI connects
DRM
-
enforcing digital audio/
video sources, such as a
set
-
top
box
, a
Blu
-
ray Disc

player, a
PC

running
Windows Vista
, a
video game console
,

or an
AV receiver
, to a compatible
digital audio

device and/or
video monitor
, such as a
digital television

(DTV). HDMI began to appear in 2006 on
HDTV

camcorders

and high
-
end
digital still cameras
.


CrossFire

is a brand name
for
ATI Technologies
' multi
-
GPU

solution, which competes with its rival
nVidia's

Scalable
Link Interface

(SLI). The technology allows a pair of
graphics cards

to be used in a single computer to improve graphics
performance. Although only recently announced for consumer level
h
ardware
, similar technology known as
AMR

has
been used for some time in professional grade cards for flight simulators and similar applications available from
Evans &
Sutherland
, ATI had also previously released a similar dual RAGE 128 consumer card called the
Fury MAXX
.


The system requires a CrossFire
-
compliant
motherboard

with a pair of
PCI Express

(PCIe) graphics cards, which can be
enabled via either hardware or software. Radeon x800s, x850s, x1800s and x1900s come in a 'CrossFire Edition' that has
'master' capability built into the hardware. One must buy a Master card, and pair it wi
th a normal card from the same
series.


Scalable Link Interface

(
SLI
) is a brand name for a multi
-
GPU

solution developed by
NVIDIA

for linking two (or more)
video cards

together to produce a single output. SLI is an application of
parallel processing

for
computer graphics
, meant
to increase the processing power available for graphics. With SLI, it is possi
ble to theoretically double the power of your
graphics solution just by adding a second video card with an identical GPU. The name SLI was first used by
3dfx

under
the full name
Scan
-
Line Interleave
, which was introduced in
1998

and used in the
Voodoo2

line of graphics accelerators.
When 3dfx collapsed financially, its intellectual property was purchased by NVIDIA. NVIDIA later reintroduced the SLI
name in
2004

and intends for

it to be used in modern computer systems based on the
PCI Express

(PCIe) bus.


The basic idea of SLI is to allow two (or more)
graphics processing units

(or GPUs) to share the work load when
rendering a 3D scene. Ideally, two identical graphics cards are installed in a motherboard that contains two
PCI
-
Express

x16 slots, set up in a
master
-
slave

configuration. Both cards are given the same part of the game (scen
e) to render, but
effectively half of the work load is sent to the slave card through a connector dubbed the SLI Bridge. For example, in
some cases the slave card will work on the bottom half of the screen. The slave then sends its rendered output to the
m
aster card, where it is incorporated into the master card's own image (in the frame buffer) and sent to the screen.



Vertex Shader


Vertices are points on a 3D map that are used to create the outlines of the images that you see within 3D games. Images
are

typically made up of many vertices, and are used to determine every object's position within the scene to be rendered.
Once each object's location has been established on the map, the map is passed to the vertex shader.

The Vertex Shader is responsible fo
r adding special effects to objects in a 3D environment. It does this by performing
mathematical calculations on the objects' vertex data using an array of variables, such as the object's co
-
ordinates, colour
and position and space. The vertex shader is re
sponsible for calculating the 3D aspects of a scene, such as colouring,
lighting etc and converting the data into a 2D map which it is passed to the pixel shader for further processing and
rendering.



ROP Unit


The Raster Operator handle the final transi
tion from the pixel pipeline to the display by building the pixel fragments
generated from the pixel pipeline into complete pixels. Most modern graphics cards have multiple ROP units. The ROP
unit also optimises the display image to save memory bandwidth,
such as when dealing with depth compression and
colour comparison.

Stream Processer


Stream processors are a relatively new technology to be introduced to graphics cards. Essentially, stream processors can
be allocated different processes to perform depend
ing on what graphical environment is to be generated. For example, in
indoor scenes the stream processors can be set as shaders, while in outdoor scenes the stream processors can be used to
map vertices.

Stream processors are commonly used in the newer gen
erations of graphics cards, replacing dedicated vertex shaders and
pixel pipelines.

The above are the main elements which make up the graphics processing power of your graphics card. There are,
however, other areas of the card which can be set and controll
ed using software such as antialiasing and anisotropic
filtering. Depending on the game you are playing, and the more powerful the graphics card you have, the higher these
additional factors can be pushed.

What is HDCP?

Short for high
-
bandwidth digital
-
co
ntent protection, a specification developed by Intel for protecting digital entertainment
content that uses the DVI interface. HDCP encrypts the transmission of digital content between the video source, or
transmitter
--

such as a computer
, DVD player or s
et
-
top

box
--

and the digital display
, or receiver
--

such as a monitor,
television or projector. HDCP is not designed to prevent copying or recording of digital content but to protect the integrity

of content as it is being transmitted.


Implementation of

HDCP requires a license obtainable from the Digital Content Protection, LLC, which then issues a set
of unique secret device keys to all authorized devices. During authentication, the receiver will only accept content once it
demonstrates knowledge of the

keys. Furthermore, to prevent eavesdropping and stealing of the data, the transmitter and
receiver will generate a shared secret value that is consistently checked throughout the transmission. Once authentication
is established, the transmitter encrypts t
he data and sends it to the receiver for decryption.

Video card

From Wikipedia, the free encyclopedia

A
video card
,
video adapter
,
graphics accelerator card
,
display adapter
, or
graphics card

is an
expansion card

whose function is to generate output images to a display. Many video cards
offer added functions, such as accelerated rendering of
3D

scenes and
2D graphics
, video capture, TV
-
tuner
adapter, MPEG
-
2/MPEG
-
4 decoding, FireWire, light pen, TV output, or the ability to connect multiple
mon
itors (
multi
-
monitor
). Other modern high performance video cards are used for more graphically
demanding purposes, such as PC games.

Video hardware can be integrated on the
motherboard
, often occurring with early machines. In this
configuration it is sometimes referred to as a
video controller

or
graphics controller
. Modern low
-
end to mid
-
range motherboards
often include a graphics chipset developed by the developer of the
northbridge

(i.e. an
nFo
rce

chipset with
nVidia

graphics or an
Intel

chipset with Intel graphics) on the motherboard. This graphics
chip usually has a sm
all quantity of embedded memory and takes some of the system's main RAM, reducing the
total RAM available. This is usually called
integrated graphics

or
on
-
board graphics
, and is low
-
performance
and undesirable for those wishing to run 3D applications. A d
edicated Graphics Card on the other hand has its
own RAM and Processor specifically for processing video images, and thus offloads this work from the CPU
and system RAM. Almost all of these motherboards allow the disabling of the integrated graphics chip i
n
BIOS
,
and have an
AGP
,
PCI
, or
PCI Express

slot for adding a higher
-
performance graphics card in place of the
integrated graphics. Despite the performance limitations, around 95% of new computers are sold with integrated
graphics processors, leaving it for the

individual user to decide whether to install a dedicated Graphics card.

History

The first IBM PC video card, which was
released with the first
IBM PC
, was developed
by
IBM

in 1981. The
MDA

(
Monochrome
Display Adapter
) could only work in text mode
representing 80 columns and 25 lines (80x25)
in the screen. It had a 4KB video memory and
just one color.
[1]

Starting with
the MDA in 1981, several video
cards were released, which are summarized in
the attached table.
[2]
[3]
[4]
[5]

VGA

was widely accep
ted, which led some
corporations such as
ATI
,
Cirrus Logic

and
S3

to work with that video card, improving its
resolution and the number of colours it used.
This developed into the
SVGA

(
Super VGA
)
standard, which reached 2 MB of video
memory and a resolution of 1024x768 at
256
color mode
.

I
n 1995 the first consumer 2D/3D cards were released, developed by
Matrox
,
Creative
, S3, ATI and othe
rs.

These video cards followed the SVGA standard, but incorporated 3D functions. In 1997,
3dfx

released the
Voodoo

graphics chip, which was more powerful compared to other consumer graphics cards,

introducing 3D
effects such as
mip mapping
,
Z
-
buffering

and
anti
-
aliasing

into the consumer market. After this card, a series of
3D video cards were released, such as
Voodoo2

from 3dfx,
TNT

and
TNT2

from
NVIDIA
. The bandwidth
requi
red by these cards was approaching the limits of the
PCI

bus capacity.
Intel

developed the
AGP


Year

Text Mode

(columns/lines)

Graphics Mode

(resolution/colors)

Memory

MDA

1981

80×25

-

4 KB

CGA

1981

80×25

640×200 / 4

16 KB

HGC

1982

80×25

720×348 / 2

64 KB

PGA

1984

80×25

640×480 / 256

320 KB

EGA

1984

80×25

640×350 / 16

256 KB

8514

1987

80×25

1024×768 / 256

-

MCGA

1987

80×25

320×200 / 256

-

VGA

1987

80×25

640×480 / 16

256 KB

SVGA

(
VBE 1.x
)

1989

80×25

800×600 / 256

512 KB

640×480+ / 256+

512 KB+

XGA

1990

80×25

1024×768 / 256

1 MB

XGA
-
2

1992

80×25

1024×768 / 65,536

2 MB

SVGA

(
VBE 3.0
)

1998

132×60

1280×1024 / 16.7M

-

(
Accelerated Graphics Port
) which solved the bottleneck between the microprocessor and the video card. From
199
9 until 2002, NVIDIA controlled the video card market (taking over 3dfx) with the
GeForce

family. The
improvements carried out at this time were focused in 3D algorithms and graphics process
or clock rate. Video
memory was also increased to improve their data rate;
DDR

technology was incorporated, improving the
capacity of video memory from 32 MB with GeForce

to 128 MB with
GeForce 4
.

Since 2003,
ATI

and
Nvidia

have domin
ated the high performance video card market with their
Radeon

and
GeForce

lines (respectively), sharing around 90% of the inde
pendent graphics card market and forcing other
manufacturers into smaller, niche markets. Most PCs computer include
Intel

integrated video, making Intel the
leading manufacturer in total volume
of video solutions, but its chipsets are not presently incorporated on
discrete video cards because of their low performance.

Components

A modern video card consists of a
printed circuit board

on which the components are mounted. These include:

Graphics processing unit (GPU)

A GPU is a dedicated processor optimized for accelerating graphics. The processor is designed specifically to
perform
floating
-
point

calculations, which are fundamental to 3D graphics rendering. The main attributes of the
GPU are the core
clock frequency
, which typically ranges from 250

MHz to 4

GHz and the number of pipelines
(
vertex

and
fragment

shaders
), which translate a 3D image characterized by vertices and lines i
nto a 2D image
formed by
pixels
.

Modern GPUs are massively parallel, and fully programmable. Their computing power is orders of magnitude
higher than that of CPUs. As consequence, they challenge

CPU in high performance computing, leading
manufacturers like Intel and AMD to integrate video, or massive parallelism, on processors.

Video BIOS

The
video BIOS

or
firmware

contains the basic program, which is usually hidden, that governs the video card's
operations and provides the instructions that allow the computer and software to interact with the card. It may
c
ontain information on the memory timing, operating speeds and voltages of the graphics processor, RAM, and
other information. It is sometimes possible to change the BIOS (e.g. to enable factory
-
locked settings for higher
performance), although this is typi
cally only done by video card overclockers and has the potential to
irreversibly damage the card.

Video memory

The memory capacity of most modern video cards
ranges from 128 MB to 4 GB. Since video memory
needs to be accessed by the GPU and the disp
lay
circuitry, it often uses special high
-
speed or multi
-
port memory, such as
VRAM
, WRAM, SGRAM, etc.
Around 2003, the video memory was typically based
on
DDR

technology. During and after that year,
manufacturers moved towards
DDR2
,
GDDR3
,
GDDR4
, and even
GDDR5

utilized most notably by the ATI Radeon HD 4870. The effective me
mory clock
rate in modern cards is generally between 400

MHz and 3.8

GHz.

Video memory may be used for storing other data as well as the screen image, such as the
Z
-
buffer
, which
man
ages the depth coordinates in
3D graphics
,
textures
,

vertex buffers
, and compiled shader programs.

Type

Memory clock rate (MHz)

Bandwidth (GB/s)

DDR

166
-

950

1.2
-

30.4

DDR2

533
-

1000

8.5
-

16

GDDR3

700
-

2400

5.6
-

156.6

GDD
R4

2000
-

3600

128
-

200

GDDR5

3400
-

5600

130
-

230

RAMDAC

The
RAMDAC
, or Random Access Memory Digital
-
to
-
Analog Converter, converts
digital signals

to
analo
g
signals

for use by a computer display that uses analog inputs such as
CRT

displays. The RAMDAC is a kind of
RAM chip that regulates the functioning of the graphics card.

Depending on the number of bits used and the
RAMDAC
-
data
-
transfer rate, the converter will be able to support different computer
-
display refresh rates.
With CRT displays, it is best to work over 75 Hz and never under 60 Hz, in order to minimize flicker.(W
ith
LCD displays, flicker is not a problem.) Due to the growing popularity of digital computer displays and the
integration of the RAMDAC onto the GPU die, it has mostly disappeared as a discrete component. All current
LCDs, plasma displays and TVs work in

the digital domain and do not require a RAMDAC. There are few
remaining legacy LCD and plasma displays that feature analog inputs (
VGA
, component,
SCART

etc.)
only
.
These require a RAMDAC, but they reconvert the analog signal back to digital before they can display it, with
the unavoidable loss of quality stemming from this digital
-
to
-
analog
-
to
-
digital
conversion.

Outputs



9
-
pin
VIVO

for
S
-
Video

(TV
-
out),
DVI

for
HDTV
, and
DE
-
15

for
VGA

outputs.

The
most common connection systems between the video card and the computer display are:

Video Graphics Array (VGA) (DE
-
15)



Video Graphics Array (VGA)

(
DE
-
15
).

Analog
-
based standard adopted

in the late 1980s designed for CRT displays, also called
VGA connector
. Some
problems of this standard are
electrical noise
,
image distortion

and
sampling err
or

evaluating pixels.

Digital Visual Interface (DVI)



Digital Visual Interface (DVI
-
I)
.

Digital
-
based standard designed for displays such as flat
-
panel displays (
LCDs
, plasma screens, wide
high
-
definition television

displays) and video projectors. In some rare cases high end CRT monitors also use
DVI. It
avoids image distortion and electrical noise, corresponding each pixel from the computer to a display pixel,
using its
native resolution
. It is worth to note th
at most manufacturers include DVI
-
I

connector, allowing(via
simple adapter) standard RGB signal output to an old CRT or LCD monitor with VGA input.

Video In Video Out (VIVO) for S
-
Video, Composite video and Component video



9 Pin Mini DIN Connector
, Frequently Used for VIVO Connections.

Included to allow the connection w
ith
televisions
,
DVD players
,
video recorders

and
video game consoles
. They often come in two 9
-
pin
Mini
-
DIN connector

variations, and
the VIVO splitter cable generally comes with either 4 connectors (S
-
Video in and out +
composite video in and out), or 6 connectors (S
-
Video in and out + component P
B

out + component P
R

out +
component Y out [also

composite out] + composite in).

High
-
Definition Multimedia Interface (HDMI)



High
-
De
finition Multimedia Interface (HDMI)

An advanced digital audio/video interconnect released in 2003 and is commonly used to connect
game consoles

and
DVD players

to a display. HDMI supports copy protection through
HDCP
.

Display
Port



DisplayPort

An advanced license
-

and royalty
-
free digital audio/video interconn
ect released in 2007.
DisplayPort

intends to
replace VGA and DVI for connecting a display to a computer.

Other types of connection systems

Composite video

Analog system with lower resolution; it uses the
RCA connector
.


Component video

It has three cables, each with RCA connector (
YC
B
C
R

for digital component,

or

YP
B
P
R

for analog component); it is used in projectors, DVD players and some
televisions.

DB13W3

An analog standard once used by
Sun Microsystems
,
SGI

and
IBM
.


DMS
-
59

A connector that provides

two
DVI

outputs on a single connector.


Motherboard interface

Main articles:
Bus (computing)

and
Expansion card

Chronologically, connection systems between video card and mot
herboard were, mainly:



S
-
100 bus
: designed in 1974 as a part of the Altair 8800, it was the first industry
-
standard bus for the
microcomputer industry.



ISA
: Introduced in 1981 by
IBM
, it became dominant in the marketplace in the 1980s. It was an 8 or 16
-
bit bus clocked at

8

MHz.



NuBus
: Used in
Macintosh II
, it was a 32
-
bit bus with an average bandwidth of 10 to 20 MB/s.



MCA
: Introduced in 1987 by IBM it was a 32
-
bit bus clocked at 10

MHz.



EISA
: Released in 1988 to compete with IBM's MCA, it was compatible with the earlier ISA bus. It was
a 32
-
bit bus clocked at 8.33

MHz.



VLB
: An
extension of ISA, it was a 32
-
bit bus clocked at 33

MHz.



PCI
: Replaced the EISA, ISA, MCA and VESA buses from 1993 onwards. PCI allowed dy
namic
connectivity between devices, avoiding the
jumpers

manual adjustments. It is a 32
-
bit bus clocked
33

MHz.



UPA
: An interconnect bus architecture introduced by
Sun Microsystems

in 1995. It had a 64
-
bit bus
clocked at 67 or 83

MHz.



USB
: Mostly used for other types of devices, but there are USB displays.



AGP
: First used in 199
7, it is a dedicated
-
to
-
graphics bus. It is a 32
-
bit bus clocked at 66

MHz.



PCI
-
X
: An extension of the PCI bus, it was introduced in 1998. It improves upon PCI by extending the
width of bus to 6
4
-
bit and the clock frequency to up to 133

MHz.



PCI Express
: Abbreviated PCIe, it is a
point to poi
nt

interface released in 2004. In 2006 provided double
the data
-
transfer rate of AGP. It should not be confused with
PCI
-
X
, an enhanced version of the original
PCI specification.



In the attac
hed table is a comparison between a selection of the features of some of those interfaces.


Bus

Width (bits)

Clock rate (MHz)

Bandwidth (MB/s)

Style

ISA XT

8

4,77

8

Parallel

ISA AT

16

8,33

16

Parallel

MCA

32

10

20

Parallel

EISA

32

8,33

32

Parallel

VES
A

32

40

160

Parallel

PCI

32
-

64

33
-

100

132
-

800

Parallel

AGP 1x

32

66

264

Parallel

AGP 2x

32

66

528

Parallel

AGP 4x

32

66

1000

Parallel

AGP 8x

32

66

2000

Parallel

PCIe x1

1

2500 / 5000

250 / 500

Serial

PCIe x4

1 × 4

2500 / 5000

1000 / 2000

Seria
l

PCIe x8

1 × 8

2500 / 5000

2000 / 4000

Serial

PCIe x16

1 × 16

2500 / 5000

4000 / 8000

Serial

PCIe x16 2.0

1 × 16

5000 / 10000

8000 / 16000

Serial

Cooling devices

Video cards may use a lot of electricity, which is converted into heat. If the heat isn't

dissipated, the video card
could overheat and be damaged. Cooling devices are incorporated to transfer the heat elsewhere. Three types of
cooling devices are commonly used on video cards:



Heat sink
: a heat sink is a passive
-
cooling device. It conducts heat away from the graphics card's core, or
memory, by using a heat
-
conductive metal (most commonly aluminum or copper); sometimes in
combination with
heat pipes
. It uses air (most common), or in extreme cooling situations, water (see
water block
), to remove the heat from the card. When air is used
, a fan is often used to increase cooling
effectiveness.



Computer fan
: an example of an active
-
cooling part. It is usually used with a heat sink. Due to the
moving parts, a fan r
equires maintenance and possible replacement. The fan speed or actual fan can be
changed for more efficient or quieter cooling.



Water block
: a water block is a heat sink suited to
use water instead of air. It is mounted on the graphics
processor and has a hollow inside. Water is pumped through the water block, transferring the heat into
the water, which is then usually cooled in a radiator. This is the most effective cooling solutio
n without
extreme modification.

Power demand

As the processing power of video cards has increased, so has their demand for electrical power. Present fast
video cards tend to consume a great deal of power. While CPU and
power supply

makers have recently moved
toward higher efficiency, power demands of GPUs have continued to rise, so the video card may be the biggest
electricity user in a computer. Although power supplies are increasi
ng their power too, the
bottleneck

is due to
the
PCI
-
Express

connection, which is limited to supplying 75 Watt
s. Modern video cards with a power
consumption over 75 Watts usually include a combination of six
-
pin (75W) or eight
-
pin (150W) sockets that
connect directly to the power supply via a
Molex connector

to supplement power.