3D graphic card info

The 3D graphic card requires much more memory than the 2D graphic card because its memory architecture is considerably different..

While a 2D graphic card only require one memory area to store ist data the 3D graphic card is using 3 specific video memory bank. The 3d graphic card memory area are named as following:

* The front buffer (required to store the image being displayed)
* The Back buffer (required to store the next image being processed)
* The Z buffer (required to store the 3rd dimension information)

Each of these buffer has to store the same color information as they are required by the 2D graphic card video memory so if for instance we were using the 16-bits color mode then each of these buffer will have to store 2 bytes for each pixel displayed on the screen for a total of 6 bytes by pixel.

Again, this can be expressed by the use a very handy formula which we will use to know how much video memory will be required for a given screen resolution and color depth under a 3D environment.

First we must determine how much bytes will have to be stored into each buffer for a given mode to know which multiplication factor we will have to include into the formula. The multiplication factor is calculated by dividing the number of bits of the color mode by 8 and this result must be added for each of the buffers

Supposing we would like to use the 24-bits color mode (16.7m colors) we will have 24/8 = 3 bytes to be stored in each of the buffers and because there are 3 buffers each storing 3 bytes each the multiplication factor will be 9 for the 24-bits color mode

now suppose we will use the 1024 X 768 screen resolution to complete the formula;

M3D = 9 X (1024 X 768) or simply M3D = 9 X 786,432

which is giving a rounded value 7.08mb, so the nearest memory size allowing this 3D color depth and screen resolution mode will be a 8mb memory bank.

Now, to be sure there will be no mistakes lets make another try and this time we will use a color depth of 8-bits (256 colors) with a screen resolution of 800 X 600.

like in the first example we will calculate the multiplication factor so we have (number of bits of the color mode = 8 which we will divide by 8 for a result = 1) then the multiplication factor will be 1 + 1 + 1 = 3 because there are 1 byte to be stored in each of the 3 memory buffers

so using the formula we will have M3D = 3 X 800 X 600 = 1.44mb which tell us that we will need at least 2mb of video memory installed on the graphic card circuit board to use the 800 X 600 screen resolution with a color depth of 8-bits under a 3D environment..

Video memory info

This memory is the one installed on the graphic card circuit board and its nature is somewhat different that the main computer memory. The graphic card memory is most of the time dual ported which means that two independent read or write operations can be accomplished simultaneously by the circuits accessing it. The graphic card components accessing the video memory are the graphic Chipset and the RAMDAC. Before the dual ported memory existed both of these component had to wait after each other before to access the memory so to overcome this problem the manufacturers developed the dual ported memory.

The dual ported memory is referred to by the name of VRAM or WRAM. The main differences between VRAM and WRAM is that the later has been specifically optimized to operate under a Windows environment so it is faster under this OS but the DOS performances of the WRAM under a DOS environment is close to the VRAM memory.

The important question about the graphic card video memory is "How much video memory is enough?".

Again, the answer to this question is very simple and can be found by the use of a small formula. The required amount of video memory needed is directly related to the screen resolution used multiplied by the color depth.

A more detailed explanation will be given here. Basically everyone knows that 1 byte = 8 bits. According to this we can now calculate that the 16-bits color mode (65k colors) will need 2 bytes to be stored into the graphic card video memory.

So, in regard of this the only thing we have to do is to first divide the color resolution number of bits by 8 to get the number of required bytes to be stored into the video memory and then multiply this value by the product of the horizontal resolution multiplied by the vertical resolution which gives the following formula;

M2D = color depth number of bits /8 X (vertical resolution X horizontal resolution) where M2D stand for memory size on a 2D graphic card.

For instance lets suppose you will use a screen resolution of 1280 X 1024 under the 16-bits color mode, then using the formula we will have:

M2D = 16/8 X (1280 X 1024) where M2D = 2.62mb so we will understand that the nearest memory size to use such a color depth and screen resolution will be 4mb because 2.62mb cant fit into a 2mb memory bank.

Note that this formula is only true in the case of a 2D graphic card.

Graphic card chipset info

The Graphic Chipset:

The graphic Chipset might be considered as the heart of the graphic card. As a fact the graphic Chipset is controlling every essential operations of the graphic card.

The way to determine the quality level of a graphic Chipset is by its maximum bandwidth. The bandwidth of a graphic Chipset is simply the width of its data path. Nowadays, there are two type of graphic Chipset in use and these are the 128-bits Chipset and the 64-bits Chipset. Like you probably guessed the 64-bits Chipset are the older one and they are in use since about 5 years. Until recently some 64-bits based Chipset were used by some graphic cards outperforming many of the newest 128-bits Chipset but now things have a tendency to equalize and as a matter of fact many of the 128-bits Chipset based graphic cards are now outperforming 64-bits Chipset which is a just return to the normality.

For many reasons some graphic Chipset are offering better performances under a Windows environment while some other are performing best under DOS environment but this cant unfortunately be established when its time to buy a graphic card unless you have some benchmark results to compare the cards you are interested by soit would be well advised to look at a large graphic card benchmark comparison listing before to go to the store so you can make sure that the graphic card you will buy is performing best under the operating system you will be using .


The RAMDAC:

earlier the RAMDAC is the circuit used to convert the digital signal of the graphic card into an analog signal that can be feed to the graphic monitor. The ability for the RAMDAC to accomplish this process is what determines its quality and this quality factor of the RAMDAC has a very important effect on the overall performance of the graphic card. The quality indices of a RAMDAC is expressed in MHZ. Before to go any further I would like to give you the terminology of these two words:

RAMDAC: stands for Random Access Memory Digital to Analog Converter

MHZ: stands for Million cycle by second

The RAMDAC frequency measured and expressed in MHZ simply means the number of operations this circuit can accomplish in a period of time of one second.

To understand this there is a very simple way and I'll explain it. Every time a new image is processed by the graphic card circuitry it has to be converted by the RAMDAC to be feed to the monitor. As a matter of fact new images are constantly feed to the monitor even if nothing changes of your monitor screen because the monitor screen needs to be refreshed at a frequency corresponding to its refresh rate but I'll come on this with more details later on this page.

Now, the amount of data that must be converted by the RAMDAC is directly related to the color resolution used by your applications multiplied by the refresh rate used plus an additional factor which has to be taken into account for the time that the cathode ray guns are tracing outside of the visible area of the monitor screen and this factor is 1.32.

According to this a formula can be defined to make this calculation like following:

MHZ = (Vertical resolution X horizontal resolution X refresh rate) X 1.32

For instance suppose we have a screen resolution of 1600 X 1200 and we will use a refresh rate of 85hz. Using the formula we will be find that a RAMDAC frequency of 1600 X 1200 X 85 X 1.32 = 215.4mhz will be required to adequately process this screen resolution at this refresh rate.

Like you can see the use of this simple formula can allow you to make sure the highest resolution and refresh rate you would like to be supported by the graphic card you intend to buy will be effectively supported by the RAMDAC installed on this graphic card.

Graphic cards

Introduction:

The graphic system is a tight coupling between the graphic card and the graphic monitor. This fact is so true that these devices should always be bought together. Graphic monitors supported performances are those allowed by its internal circuitry and the included cathode ray tube as well as the supported performances of the graphiccard driving it.

A good monitor is by definition a monitor with sufficient displaying surface to allow a comfortable readability without effort from the user, a close matching to the real natural colors, enough brightness and contrast to give realistic details even when the images are not ideal, no visible flicking so it wont tear your eyes even after a long period of use, a choice of color resolution large enough to allow an interesting variety of color resolutions and a screen dot pitch small enough so the images details will look close to photo realistic quality. In numbers, this would mean at least a 15 inches monitor capable to support a refresh rate of at least 72hz to 75hz at a screen resolution of 1024 X 768 with a dot pitch of .28 which actually is the case of most of the middle grade monitors. For better performances a 17 inches monitor is more flexible but in this case you should choose one that would support screen resolution of 1600 X 1200 at a refresh rate of at least 85hz at the highest screen resolution with a dot pitch of .25 so the images wont look grainy while no flicker will be perceptible.

A dot is the smallest possible portion of a screen monitor and its size is varying according to the picture tube quality while a pixel is a determined area covering a portion on the monitor screen which size is determined by the application is use. In other word a pixel can use more than one screen dot but can never be smaller than one screen dot. However, on a high definition monitor a pixel can be so small on the monitor display that if someone would use a black background screen where only a single white pixel would be displayed it could be very difficult to locate it by most of the people.

To summarize all this we could say that a graphic monitor quality can be determined by the following criteria:

* A small dot pitch
* A high screen resolution
* A high refresh rate
* A large diplay surface

The relative importance of these criteria are very much depending of your own personal needs and budget of course. The smaller the dot pitch or the largest the display surface of a monitor the more expensive the monitor usually is so you may have to sacrifice some display surface to get a monitor size that fits your budget. A good rule is to get the smallest possible screen dot pitch that you can afford for the monitor size you intend to buy. In any case you should never change an existing monitor for a larger one using the same dot pitch if you don't want to lose any image definition because the large the display surface is the lower its definition will be if its dot pitch is not reduced accordingly. For instance, on very large display you will have to establish a good distance between you and the monitor if you don't want to see the granularity effect. While this effect is more perceptible on TV screen it can also be observed on computer graphic monitor display using very large screen.

Just as a reminder I would like to summarize the different color depth and their corresponding names so you could use it in the suite of this page where these names will be widely referred to:

* 8-bits = 256 color named the VGA mode
* 16-bits = 65,536 colors named the high color mode or SVGA
* 24-bits = 16.7m colors named True color mode
* 32-bits = 16.7m colors named True color mode



Graphic cards Basic Hardware informations:

A graphic card is a very complex device. As a matter of fact there are so many components on a graphic card circuit board that it could almost be considered as a small stand alone computer. Like on the computer motherboard the graphic card is using its own clock circuitry to time each of the computed operations it perform, the graphic card also possess its own BIOS which like the computer motherboard can be software upgradable (Flash BIOS) or not and the graphic card circuit board also has its own self contained Chipset which in this case is referred to as the graphic Chipset and finally the graphic card also uses its own memory sockets. There are also some specific circuitry like the RAMDAC which is used to convert the digital signal form to an analog signal form as required by the graphic monitor as well as the expansion connector to allow the use of add-on card for additional feature like Mpeg acceleration, TV-Tuner, ect, ect..

The goal of this page is not to review every technical aspects of the graphic card but only the most important aspect that can be useful when its time to choose a graphic card. So, according to this I will try to mainly focus on these topics and give as much details as I can without being too technical.

These are in my opinion the most important criteria to consider when choosing a graphic card:

* The RAMDAC frequency
* The highest possible resolution
* The highest possible refresh rate
* The maximum memory allowed

Now we will have a quick look at some of the most important component used on graphic cards and we will try to see what are their importance on the overall performances of the graphic card.

Hardware intro:

Introduction:

Hardware is the physical manufactured part of a computer. The most visible example of hardware is the casing of the computer. Any physical part of a computer is hardware. These include the motherboard, CPU, ram, harddisk, floppy disk, keyboard, mouse and monitor. Other hardware items that are often mentioned include printers, modems, CD and DVD drives, scanners and other peripherals.

Hardware items require a to work properly or to fully utilize their capabilities. The device driver is needed to act as a bridge between the operating system and the device. This is so user level programs don't have to worry about interacting with different devices. The device driver provides a standard interface independent of the device.

Examples of hardware that requires device drivers include video cards, sound chips, and Ethernet interfaces. Usually the required driver or drivers are included on a CD-ROM with the item, but sometimes it is necessary to get the latest version from the Internet.

Hardware is often categorized according to its interface to other devices. Common interfaces include USB (universal serial bus), firewire (IEEE 1394), PCI bus, RS-232 serial, and parallel port.

Computer:

A computer is defined as any device which computes. The word comes from Latin meaning "to work out together with". The first use of the word computer did not apply to devices but to people. The word computer was first applied to a mechanical device in 1897.

In modern times a computer is an electronic device where data is input via various means. The data is manipulated via some programs(instructions), and an output is produced either electronically or physically.

Electronically output data can be in the form of files or screen display. Physically output data can be in the form of printed paper or physical movements such as motor control of robots. A computer is available in various sizes, configurations and consists of hardware and software.

Computers can be divided into different modules that when combined work together as a computer. They consist of a computing unit which processes a program. A memory unit which provides a working space for the computing units calculations. A storage unit for long term storage of data. An input unit for inputting programs and data for the computer to work on and finally an output unit for the result of the calculations. The first computing device which had all of these modules integrated as one machine was the Analytical Difference Engine invented by Charles Babbage in 1834. This was a mechanical device which never actually worked but is considered the forerunner of modern computers which use the principles of Babbage's machine.

Computers can be use as stand-alone or in a network. A network can be business or home. Even 2 computers can form a network.

Clustering(harddrives)

The term "cluster" can refer to the way a magnetic disk (hard disk drive or floppy drive) is logically divided, in order to organize and manage the storage area or the concept of "grid computing"

It is also encountered as a measurement in hard drive partitioning and allows a single hard drive to be partitioned to support different file structures.

Partitions are when a drive is broken up into different portions. Each portion, or partition, has its own boot record and FAT.

Magnetic Storage Area

A cluster is the logical unit of file storage on a hard disk and is managed by the computer's operating system. Clusters help to organize and to identify the disks content. Clusters, generally used for increased speed and accuracy, come with individual ways to be identified so that they are easily traceable on a hard disk. A cluster is also used to group together machines in order to give the appearance of being one machine because they are performing tasks that closely parallel with one another.

On a hard disk drive, a cluster, when taken with the file system (i.e.: FAT32, NTFS etc.) will determine the total number of files, the total capacity of the drive and the efficiency of storage on the system.

Grid Computing

A number of small, low powered, computers can be a group of computers connected together so that they can work together like they are one. This is the essence of "grid computing."

Partitions and Clusters - FAT, NTFS and HPFS

FAT stands for "File Allocation Table" and systems with this designation had a fairly simple structure with minimal space allowed to address the cluster allocations.

NTFS was designed for Windows NT and introduced many advanced features like data encryption and cluster compression. In today's computing world many of the features of our modern file systems would be absent without an NTFS file system. The main feature missing would be security, down to individual file locking.

FAT12 FAT12 uses only 12 bits for cluster addressing. FAT16 came soon afterwards.

FAT16FAT16 uses 16bit locations to reference files. This limits how many files there can be. There could be only 65536 files (because this is all that 216 clusters will allow!) FAT16 cluster size depends on the size of the hard disk. Because of the limitation of files the larger hard drives would waste space because the smallest 1 byte file would need a 32k cluster to be in the system! And no file could be larger than 2 Gbytes. Therefore a hard drives which is more than 2 GB's would have to be partitioned into multiple drives. (What would be do with the current 160 Gbyte hard drives on a FAT16 system? Not much!)

FAT32 and VFAT (Virtual File Allocation TableFAT32 uses 32bit locations to reference files. This file system came in with Win95. With FAT32 you could address 4294967296 clusters, or use a hard drive capacity of 2 terabytes. (It is truly amazing how advanced Win95 was.

New Technology File System(NTFS) is used exclusively with Windows NT and Windows NT based Operating Systems. It uses 32-bit locations to reference files like FAT32/VFAT, but data encryption and cluster compression is also possible. This is why these files systems are so compatible with each other. This file system was introduced by Microsoft Engineer Helen Custer. Helen Custer spent three years as part of the Windows NT design team.

HPFS (High Performance File System) HPFS first appeared is IBM's OS/2 Operating system. It is no longer used by any other OS, but it did feature support for 2TB disks and filenames up to 256 characters.

History

When IBM and Microsoft were both developing the alternative operating system, OS/2, they realized that the File Allocation Table system, FAT, was not a useful system for hard disks, and Microsoft began developing the High Performance File System, codenamed "Pinball."

Instead of coding it inside the kernel, as FAT was, Microsoft developed a "driver-based" Filesystem API that could allow them and other developers add new filesystems to the kernel without needing to modify it.

When Microsoft stopped working on OS/2, IBM continued using the IFS interface and Microsoft implemented a similar one in their Windows NT product.

Data transmission

Wireless devices:
A wireless device is a device which transmits the data or information through radio or electromagnetic waves, and does not require any physical connection, like a copper or optical fiber cable. A wireless network connection is required for the devices to work.

Peripherals:
A peripheral is a piece of hardware that hooks up to your computer. Peripheral devices can be hooked directly to the computer or it can be wireless. Some popular peripherals are, printers, hard drives, plotters, etc.

Sometimes these peripheral devices in order to operate will require the installation of drivers. These usually come with the item, but sometimes it is necessary to get the latest version from the Internet.

Peripherals are also labeled according the "hookup" type. That is the type of cable required to hook it up to the computer. The more popular connections these days are USB, Firewire and Wireless. Some other connection types from the past are SCSI, Serial and Parallel. While these connections still exist they are usually used only on older machines.

A peripheral device can also be a stand-alone unit independent of the computer. These include, but not limited to, devices like magnetic tape cleaners and/or testers.