Note: Periodic Updates will be made to this guide as needed.

Last Update: 06-27-2006

A lot of us usually hear some fancy hardware terms used while talking about our PC components as well as their functions and parts , which are very intriguing but we actually don’t know what they EXACTLY mean and REALLY do . This guide is written to clear up some of these confusions about your PC hardware .

Please Note that the links to the pictures included in this guide are from several different websites . These links are posted so that you also get to visually KNOW what you are reading .

The Processor :

The exact term processor is a sub-system of a data processing system which processes received information after it has been encoded into data by the input sub-system. These data are then processed by the processing sub-system before being sent to the output sub-system where they are decoded back into information.

In other words it’s the brain of your system controlling everything . So the faster your processor speed, the faster will be your overall system and vice versa ( Of course your other components are also involved in the overall effective speed of your PC at any task ) .

A Simple Processor [ Pictures ] :

From Above :



From Below :



Different processors have different sockets e.g. AMD processors come in sockets A, 754 , 939 and 940 . Intel has 478 , 775 sockets (also known as LGA) etc . so you will have to check which socket your motherboard supports before getting a new processor .

Cache :

Also called cache memory . Cache (computer), in computer science, an area of memory that holds frequently accessed data or program instructions for the purpose of speeding a computer system's performance. A cache consists of ultra fast static random-access memory (SRAM) chips, which rapidly move data to the central processing unit. The process minimizes the amount of time the processor must be idle while it waits for data. This time is measured in a clock cycle, which is the equivalent in time to a bit in data. The effectiveness of the cache is dependent on the speed of the chips and the quality of the algorithm that determines which data is most likely to be requested by the processor .

L1 Cache :

The primary cache build into the CPU . The memory closest to the CPU as well as the smallest of the all three caches .

L2 and L3 Cache :

(Level 2 cache) A memory cache that is external to the CPU chip. Typically located on the motherboard, the L2 cache is also known as a "secondary cache." An L2 cache feeds the L1 cache. If the L2 cache is also contained on the CPU chip, then the external motherboard cache becomes the L3 cache. The L3 cache feeds the L2 cache, which feeds the L1 cache, which feeds the CPU.

The Chipset :

A set of chips that provides the interfaces between all of the PC's subsystems. It provides the buses and electronics to allow the CPU, memory and input/output devices to interact. PC chipsets, which are housed on one to four chips, include built-in controllers for almost all common peripherals. The primary components on a PC motherboard are the PC chipset, CPU, memory, clock, buses and BIOS .

The Northbridge/Southbridge chipset architecture has been widely used in both Intel and AMD PCs. The Northbridge part handles high-speed channels, while Southbridge manages lower-speed channel .

Bus and Front Side Bus ( FSB ) :

In easy language it’s the primary pathway between the CPU and peripherals .It’s speed is determined from the number of parallel channels (16 bit, 32 bit, etc.) and clock speed. The Front side bus is faster than the peripheral bus (PCI, ISA, etc.), but slower than the backside bus . This bus carries all information that passes from the CPU to other devices within the system such as RAM, PCI expansion cards, hard disks, etc. It should be noted that you may read about terms Real and effective FSB . The real FSB is the actual speed of the FSB and effective FSB is the overall effect of that speed . This varies on different brands of processors i-e on Intel the FSB can take four time the data of one standard cycle so giving you an effect of 800Mhz by actually running at 200Mhz . Similarly , on AMD the FSB can take twice the data of one standard cycle and give an overall effect of 400Mhz from 200Mhz actual speed .

Several chip makers are looking to dual-core designs as a way to increase the performance of next-generation processors without the constraints imposed by rising levels of power consumption in single-core processors. Two common methods of improving the performance of a single-core chip are increasing clock speed or adding cache memory, both of which require more power.

A dual-core chip is basically two separate processors on a single chip. Those two processors can outperform single-core processors on most multithreaded applications while running at lower clock speeds and consuming less power.

An application with multiple software threads will run faster on a dual-core processor because the operating system can assign an individual thread to its own processor core. Multithreaded applications running on a single-core processor must wait for one thread to finish before another thread can be processed.

Basically, Dual Core will help you hear mp3s in media player while browsing websites and installing softwares, all at the same time. By doing so, it gives you the power of multitasking more efficiently, a task at which single core CPUs usually made you frown due to their slow performance.

Note: It doesnt NECCESARILY MEAN that if you have a dual core processor with a multitasking Operating System, you are going to get blazing performance at every application out there. To some extent, the application developer must have also OPTIMIZED his application for Dual Core Processors. Only that way you can feel the true potential of Dual Core. Currently, all AMD all latest AM2 and 939 socket processors (with X2 in the title) and some of Intel's Pentium Processors toghether with Intel's Core 2 duo "Conroe" are dual core. 

Motherboard :

Its also known as main board , The motherboard is a printed circuit board that connects the various elements on it through the use of traces, or electrical pathways. The motherboard is indispensable to the computer and provides the main computing capability .

Your motherboard should look something like this :



Motherboards also contain important computing components, such as the basic input/output system (BIOS), which contains the basic set of instructions required to control the computer when it is first turned on; different types of memory chips such as random access memory (RAM) and cache memory; mouse, keyboard, and monitor control circuitry; and logic chips that control various parts of the computer’s function. Having as many of the key components of the computer as possible on the motherboard improves the speed and operation of the computer .

Users may expand their computer’s capability by inserting an expansion board into special expansion slots on the motherboard. Expansion slots are standard with nearly all personal computers and offer faster speed, better graphics capabilities, communication capability with other computers, and audio and video capabilities. Expansion slots come in either half or full size, and can transfer 8 or 16 bits (the smallest units of information that a computer can process) at a time, respectively.

( Nota bene : Expansion slots are a rare sight in modern motherboards , they used to exist more commonly in previous generation motherboards ) .

The pathways that carry data on the motherboard are called buses. The amount of data that can be transmitted at one time between a device, such as a printer or monitor, and the CPU affects the speed at which programs run. For this reason, buses are designed to carry as much data as possible. To work properly, expansion boards must conform to bus standards such as integrated drive electronics (IDE), Extended Industry Standard Architecture (EISA), or small computer system interface (SCSI).

The motherboard also has different slots for different devices . Like the old PCI slot (Peripheral Component Interconnect ) Its one of the most commonly used buses ( this is the slot in which your dial up modems , sound cards , previous 3D cards as well as the T.V. tuner cards ) . Next to it comes the AGP slot/bus ( Accelerated graphic port ) used only to plug in your 3D card ( as it replaced the PCI slot , at least for 3D cards ). AGP came in 3 flavors i-e AGP 2x , 4x , 8x . The primary difference between them is bandwidth increase ( the higher the bandwidth the faster the link of the CPU with the card , hence better performance but cards up till now don’t fully use such increased bandwidth so the difference from AGP 4x to AGP 8x is only 1Fps ) . AGP is also being replaced by a new slot called PCI Express . It supports a bandwidth up to 16x and gives us the ability to use two 3D cards simultaneously in our motherboards to have an increase up to 1.9x the processing power of the previous card (in certain situations). This technology is introduced both by ATI and NVIDIA as Crossfire and SLI respectively .

Motherboard Manufacturers are heavily advertising SLI capability of their Motherboards to boost Sales :

RAM :

RAM is an acronym for Random Access Memory. Semiconductor-based memory that can be read and written by the microprocessor or other hardware devices .

A typical RAM module :



SDRAM :

(Synchronous DRAM) A type of DRAM memory chip that has been widely used since the latter part of the 1990s. SDRAM chips eliminate wait states because they are fast enough to be synchronized with the CPU's clock. The chip is divided into two cell blocks, and data are interleaved between the two. While a bit in one block is accessed, a bit in the other is prepared for access. This allows SDRAM to burst subsequent, contiguous characters at a much faster rate than the first character.

DDR and DDR2 :

Double Data Rate SDRAM (DDR) doubles transfer rates by transferring data on both the rising and falling edges of the clock so a PC3200 ( DDR 400 ) ram is actually operating at 200MHz while giving the speed of 400 MHz . DDR uses additional power and ground lines and is packaged on a 184-pin DIMM module rather than the 168-pin DIMM used by the first SDRAM chips. Laptops use 200-pin SODIMM modules.

DDR2 chips increase data rates using various techniques such as on-die termination, which places the terminating transistors that eliminate excess signal noise on the chip itself. DDR2 modules require 240-pin DIMM slots, and although they are the same length as DDR, they are keyed differently and will not fit into the DDR slot. DDR 2 modules are known to produce quite much heat too so their latency is gradually increased over a long period of operation and thus their performance decreases .

RD Ram :

RD RAM is short for Rambus DRAM . It supported higher speeds than DDR or SDRAM but are rarely seen in CPUs these days , although they were supported by Intel .They are used in Nintendo 64 and PS2 and they are quite expensive . Rambus's successor to RDRAM is XDR DRAM, used by IBM’s "Cell" processor technology and the PlayStation 3.

Dual Channel DDR and DDR2 :

Chipsets on the motherboard may support two independent memory controllers, which allow access to two memory modules simultaneously (upstream data on one 64-bit channel; downstream data on the other). Channels can be configured as two 64-bit or one 128-bit. Modules must be installed in matched pairs unless the chipset has an option for mismatched modules . Motherboard manufacturers often make the slots ( which support Dual Channel DDR modules ) with the same colors so that they can be easily recognized :



Memory Timings and FSB Relation :

This a quite difficult concept to comprehend and you don’t need to know much about these timings in detail so for the average Joe , I’ll just say :

With all things equal, a stick of DDR memory capable of running 2-3-3-6 will make the computer operating experience seem faster than a DIMM which may only run at 3-4-4-8. This is because the delay from when the memory receives an instruction, retrieves the data, and sends it back out is less.

But I know some of you will really want to dig into this stuff so for you guys , HERE WE GO :

The memory timings shows you the latencies .The memory timings of any
module is often written in the following format :

2-3-3-6-1T

The first digit i-e 2 , denotes CAS latency (CAS being an acronym for column address strobe) a signal, or strobe, sent by the processor to a DRAM circuit to activate a column address. DRAM stores data in a series of rows and columns, similar in theory to a spreadsheet, and each cell where a data bit is stored exists in both a row and a column. A processor uses CAS and RAS (row address strobe) signals to retrieve data from DRAM. When data is needed, the processor activates the RAS line to specify the row where the data is needed, and then activates the CAS line to specify the column. Combined, the two signals locate the data stored in DRAM . The CAS Latency (CL) refers to the length of time, in clock cycles, it takes for a request sent from the memory controller to read a memory location and send it to the module's output pins .

Next in line is RAS-to-CAS Delay , in this case the second digit i.e 3 . This field allows you to set the number of cycles for a timing delay between the CAS and RAS strobe signals, used when DRAM is written to, read from or refreshed. Lower settings result in faster performance. 3T, 2TBank Interleave .

This number represents the term , TRP which indicates how fast RAM can terminate one row access and start another one .

Next is Active to Precharge (tRAS) which controls the length of the delay between the activation and precharge commands -- basically how long after activation can the access cycle be started again . ( here its 6 )

Lastly 1T / 2T thing ( note T denotes tick ) , I will quote this since there could not be a better and easier definition, “ BIOS feature controls how long the memory controller latches on and asserts the command bus. The lower the value, the faster the memory controller can send commands out .”

There is a definite proportion between the increase / decrease of FSB and ram speed in MHz . Simply saying that when you increase your FSB your RAM speed also increases automatically but in a specific ratio which in turn depends on the stability of the Ram / FSB over clock . The ratio is called “Memory Divider”.

Hard Drive :

Hard Disk, in computer science, one or more inflexible platters coated with material that allows the magnetic recording of computer data. Hard disks provide faster access to data than floppy disks and are capable of storing much more information. Because platters are rigid, they can be stacked so that one hard-disk drive can access more than one platter. Most hard disks have from two to eight platters.


THE INSIDE OF YOUR HDD :



The inside of a computer hard disk drive consists of four main components. The round disk platter is usually made of aluminum, glass, or ceramic and is coated with a magnetic media that contains all the data stored on the hard drive. The yellow armlike device that extends over the disk platter is known as the head arm and is the device that reads the information off of the disk platter. The head arm is attached to the head actuator, which controls the head arm. Not shown is the chassis which encases and holds all the hard disk drive components.

HDD come in speeds of 5400 RPM ( Rotations per minute ) 7200 RPM as well as 10000 RPM . Currently the 7200 RPM HDDs are the standard while 10000 RPM are quite expensive to buy . 15000 RPM drives have also recently popped up. Although they offer ultra fast performance, they will also take a very big chunk out of your bank account. The greater the speed of the HDD the lesser time it will take to LOAD games and programs .

PATA AND SATA HDDs :

(Parallel ATA) Refers to the original ATA (IDE) technology that uses a parallel data channel from the controller to the disk drives. After Serial ATA drives became popular, the PATA term was coined to specifically refer to the parallel drives.

( Serial ATA ) utilizes new techniques to increase the rate of data transfer between HDDs ( upto 1.5GB/sec) . In these HDDs no HDD acts as a master or a slave HDD . In fact both are connected to their interface . The connectors of SATA HDDs are also considerably smaller than PATA HDDs . Therefore , your CPU doesn’t look much crowded :

An Image showing Serial and Parallel ATA :



SATA II HDDs:

Recently SATA II HDDs are revealed . They have much higher speeds ( 3.0GB/sec ) . New Nforce4 chipset based motherboards like ASUS A8N-SLI DELUXE and Giga-byte GA-K8NXP-SLI that support 3.0GB/sec SATA transfer rates can use this new technology. Since, now due to wide availability, the prices have now dropped within the range of the average joe to get advantage of it so i'll recommend all of you to go for them instead of SATA I HDDs.

3D Cards :

The Cores of our Gaming rigs are our 3D cards . Most of us have often heard the names of ATI’s Radeon line of 3D cards as well as Nvidia’s Geforce series . Now we shall know about what they do and how they do it .

A Typical 3D Card :



A lil’ History :

The pioneers of 3D cards were the 3D FX who introduced some revolutionary technologies in Gaming . They introduced their first Voodoo card in 1996 and continued the line up to Voodoo 5 6000 ( which never saw the day light ) . They introduced techniques like Anti Aliasing and SLI ( Scan Line Interleave ) . In late 2000, several of 3dfx's creditors decided to initiate bankruptcy proceedings. 3dfx would have had virtually no chance of winning these proceedings, and instead opted to be bought by NVIDIA, ceasing to exist as a company. Most of the design team that were working on Rampage (the successor to the VSA-100 line) were transferred to the team working on what has since become the GeForce FX series. The Rampage 3D card , which was in development , thus never came out .

HOW ARE 3D CARDS MANUFACTURED :

The 3D – cards are first designed by companies like ATI and NVIDIA . Within the company first talks are continuously held between marketing department and engineers . Finally a budget and a manufacturing process is decided upon ( e.g. 130nm , 110nm or 90 nm ) . Later on , the chips are then designed around HDLs ( Hardware Description Languages ) . The companies then thoroughly tests their cards for heat and compatibility issues and after ensuring manufacture some reference cards ( a.k.a. reference samples ) . These cards are sent to NVIDIA/ATI launch partners with their card chipsets so that they can be used to make the cards . These companies e.g. ASUS , MSI , BFG etc then make their own cards by following the design of the reference card ( off course some companies customize their cards so as to get an edge over their rival companies e.g. BFG always release an Over clocked version of their card ) . These cards are then brought to retailers and etailors for sale .

THEIR ARCHITECTURE :

A typical fairly modern 3D card will have the following features :

Pixel Shader Pipelines supporting a specific pixel shader version e.g Geforce 7950 supports Pixel Shader 3.0 while Radeon X700 supports Pixel Shader 2.0b. They also have specific Vertex Shader pipelines which support a specific Vertex Shader Version . This architecture is made under Microsoft direct X specification ( API ) . Microsoft next generation of API will soon replace
Direct X 9 with its version 10 which will support Shader Model 4.0 with the launch of Windows Vista (currently scheduled for early 2007 release). This new version is supposed to require the card to have a new Unified Shader Architecture i-e cards will have unified shader pipes with vertex and pixel shader units ( but this is the future so we will see soon what next generation will bring us ) .

Around 128/256/512/1GB VRAM , built into the card itself . This VRAM ( Video Random access Memory ) is specially available to card to swap data with the rest of the system .

AGP/PCI-E slot support ( in case of PCI-E newer cards also support SLI )

A large no. of transistors ( most probably in millions) to give the card enough RAW power .

128/256 bit memory interface ( the wider the memory bus interface , the better the card will perform ) and a Fill rate ( in thousands )

( Nota bene : From a buying point of view you have to consider many other things like price/real time games performance , cooling solution , warranty as well as the product’s support by game developers ) .

The data in the GPU is processed in many ways . DirectX or OpenGL commands and HLSL and GLSL shaders are translated and compiled for the architectures. Commands and data are sent to the hardware where we go from numbers, instructions and artwork to a rendered frame.

The first major stop along the way is the vertex engine where geometry is processed. Vertices can be manipulated using math and texture data, and the output of the vertex pipelines is passed on down the line to the fragment (or pixel) engine. Here, every pixel on the screen is processed based on input from the vertex engine. After the pixels have been processed for all the geometry, the final scene must be assembled based on color and z data generated for each pixel. Anti-aliasing and blending are done into the framebuffer for final render output in what NVIDIA calls the render output pipeline (ROP) .

All of this basically doesn’t concern the end user as he only needs to know about real time performance so I’ll just tell something about each feature to make you know the worth of it .

VRAM : Now-a-days 64 MB cards are not recommended and newer cards don’t even have 64 MB counterparts . For mid range cards 128 MB/256 MB ram should suffice . High end users should keep their eye on 512 MB versions as their beasts ( X1900 / 7900 GTX etc .) usually need high amount of VRAM to utilize their capabilities . And if you have even more cash to shell out for your graphic card, you can get the 1GB VRAM Nvidia's 7950 GX2 monster, which has basically two 7900GT cores. 2 of these babies can also give you the capability to go Quad Sli.

( Note : Quite recently , 512 MB/1GB cards have also appeared . These cards have very high clocks and they can utilize these additional VRAM. It is still mostly unclear whether the increase in performance is due to this VRAM or their other architectural Capabilities. )

SLOT : For the slot , PCI-E has quite an edge over AGP . First of all , most of the 3D card manufacturers have started to move from AGP to PCI-E so PCI-E is future proof . PCI-E motherboards are also getting cheaper now . PCI-E offers twice the bandwidth of AGP ( 16x PCI-E versus 8x AGP ) . SLI/Crossfire are exclusive to PCI-E as well . PCI Express is a point-to-point connection, meaning it does not share bandwidth but communicates directly with devices via a switch that directs data flow. The initial rollout of PCI-Express provides three consumer flavors: x1, x2, and x16. The number represents the number of lanes: x1 has 1 lane; x2 has 2 lanes, and so on. Each lane is bi-directional and consists of 4 pins. Lanes have a delivery transfer rate of 250 MB/ps in each direction for a total of 500 MB/ps, per lane . PCI-E 1x and 2x slots will be used for simple devices while PCI-E 16x is used for 3D cards .

( Nota Bene : PCI Express is not to be confused with PCI-X, used in the server market. PCI-X improves on standard PCI bus to deliver a maximum bandwidth of 1GB/ps. PCIe has been developed for the server market as well, initially with the x4, x8 and x12 formats reserved. This far exceeds PCI-X capability. As graphic demands increase, x32 and x64 slots will be realized, and future versions of PCIe are expected to drastically increase lane data rates. )

Transistor Count : This determines the RAW POWER of a graphics card . This can also be judged by running a game at minimum graphical settings while taking into account the FPS . This shouldn’t bug you in any way .

128/256/512 Bit memory interface and Fill Rate : These determine the speed at which the card can process the data ( Of course this is also determined by the memory clock speed ) so the higher the better . 128/256 Bit memory interface cards are used mainly in mid-range gaming cards . Most of Nvidia's High End cards are 256 bit but they give very high performance over their competitors due to superior architecture and better driver support, but ATi has also unveiled some 512bit cards that bring some zest into the high end segment.

Manufacturing process : It helps the Card manufacturers determine the amount of transistors packed on the card and thus contributes to the cost of a card . Hence , it shouldn’t concern the end user .

--------------------------------------------

This concludes the guide . Hope it helps you understand your PC better .

( P.S : Nearly every definition/Explanation was confirmed from various sites to ensure accuracy as well as to cover the most recent in technology . )