Thursday, July 29, 2010

PCI, PCI-X, PCI Express - Comparison

A brief intro on Buses and Peripherals

look like "a bundle of wires" or PCB's that terminate at multiple connectors where devices can be plugged in. They are used basically to connect electric components.

Early computer buses were just bundles of wire connecting memory with peripherals(External devices/Hardware which need the host perform a function eg: Mouse, Keyboard.)

The S-100 bus "Card Cage" - The S-100 bus was used in the earliest "Microcomputers"

Technology evolved, computers became sophisticated, and so did these guys.

What we use today:

PCI (Peripheral Component Interconnect)

PCI is a widely used computer bus which acts as a standard interface for connecting peripherals to the computer, in two ways - An integrated circuit embedded in the motherboard, or an expansion card which fits into it's slot.

PCI is capable of allowing high speed of data transfer, which modern peripherals(eg: Graphic cards, Ethernet)require.

More info in PCI found here:


PCI-X(64 bit) is the 32 bit PCI bus optimized for higher performance. It is twice the width of the PCI and narrower. It runs at higher frequencies, and can manage four times the clock speed. It was developed(by IBM, Microsoft, HP) to increase the performance of high level peripherals like high-end graphic cards and Storage devices, gigabit - speed ethernet, which isn't feasible by the conventional PCI design.

Except for a few changes, The basic structure of PCI and PCI-X are similar, they follow the same protocol and perform the same functions.

PCI Express/PCIE

PCI Express is an innovative new design that is used as an Expansion Card(It can't be connected to PCI/PCI-X directly) to replace PCI and PCI-X.

PCIE operates more like a network than a bus.

In a PCI/PCI-X, one bus handles data from multiple sources at once.

In PCIE, a switch controls several point to point serial connections, these connections are conveyed directly to the device where the data needs to go making a serial interconnect along a switched bus, dedicated exclusively to that slot. Meaning - The PCIE point to point connection connects only two devices at one time.

Each PCIE slot is attached to the motherboard using a unique "lane", which does not share the data path with other PCIE Slots. Multiplying individual data "lanes" to produce interconnects which deliver upto 16 times the bandwidth of a single lane.(This is the reason their slots are referred to as PCI Express*4/PciExpress*16)

Further information on how PCI Express works:

PCIE vs PCI-X - Why PCIE is Superior

Parallel connections (used by PCI-X) undergo problems during high speed data transfer - like electromagnetic interference.
Interconnecting lanes of PCIE give it a higher bandwidth.
Less steps needed to process instructions, due to it's architecture.
Add more.

Performance Report: 3GIO stands for third generation input-output, another name for PCIE
BW -Bandwidth, Pin-Number of Pins

Statistics and data on PCI, PCI-X, PCI Express found here:

How to differentiate b/w the three:

Pictures pulled from:

If any you feel anything mentioned is incorrect, please drop a line and I'll make changes to ensure no misconceptions happen.

File size limit of 4GB in FAT32 file system

Many of u may not be familiar with this that there is a file size limit of 4GB in file system( or more technically its 3.99 GB)... Even i am a years old FAT32 system user yet i had no idea abt it till recently!!!

There are two well explained reason behind not knowing abt it, first being that most of us have switched to the modern file systems such as exFAT/FAT64, NTFS or NTFS5... Second reason being we pretty rarely come across with the files larger than 4GB, and even if we come across them the are split into volumes of smaller sizes,so we may not know abt them... And the last and most common reason we actually never cared!!!

Reason for the file size limit lies in the design itself the FAT32 system was designed on the basic FAT systems and took it to a much higher limit from the parent version.. This file system's maximum limiting size on stored files is set by its inherent design considerations. The FAT32 file size limit is only 4 GB minus 1 byte (232 - 1 bytes). This limitation was imposed because with greater volume sizes, data transfers becomes too slow under FAT32 format..
Hence in that case Microsoft wanted us to use the other advanced file systems...

The newer systems provided better security features, better file management, stability, speed, support for large hard disks, compression and above all about unlimited file size and cluster size... The latest of the file systems is NTFS5, which is a slight improvement from the previous version of NTFS, more about this will be in my next blog.. Till then u'll have to wait!!!

Here is a picture of the files larger than 4GB and how they are splitted in parts of 3.99GB in a FAT32 system:

Here's a list of file systems and their file size limits:

  • FAT12: 16MB
  • FAT16: 2GB
  • FAT32: 4GB minus 1 byte
  • exFAT/FAT64: 16ZB
  • NTFS: 16TB minus 64KB
  • NTFS5: 16 ExaBytes minus 1KB
sources: Microsoft support center

As this was my first blog, i apologize if i made a mistake in writting anything, and suggestions are alwayz welcomed... As of the blog i understand its too small, but i have been typing and posting everything through cellphone(since i m having a bandwidth problem), n u all must know how irritating its to type on a cellphone, yet i have tried to make the post a better one... Hope u like it...

Hard Disks: Advantage of SATA over IDE

Hi everyone! I had asked a question in SM Lab Session (I don't even exactly remember the question now!) And I was told to 'Google it & Blog it'. So here I am blogging about IDE & SATA: Advantage of SATA over IDE.

Every hard disk is interfaced in the motherboard through a disk controller. This disk controller is the one that plays bridge between the hard disk and the operating system. For the user to access data from the hard disk, there is a data bus connecting the hard disk and the motherboard. The data bus is like a pipe where bytes of data are being transmitted.

There are two kinds of data buses, Serial ATA and Parallel ATA, which are used in SATA hard disk and IDE hard disk respectively.
IDE is the abbreviation of Integrated Device Electronics. (The Parallel ATA was earlier called IDE. That
s why hard disks that use Parallel ATA are called IDE hard disks.)

SATA stands for Serial Advanced Technology Attachment.

The concept of the SATA was ratified by the ANSI (American National Standards Institute) in year of 2002.


  • SATA has the capability to provide point to point channel communication between mother board and the drive. Where as in the IDE, the architecture of the master slave combination is supported on a single cable for the two drives it supports. (In the case of IDE, the controller allows the connection of 2 peripherals. The band-width is thus divided between the 2. This reduces the performances. On the other hand, in SATA, each hard disk is connected to its own controller via his own cable. This implies that there are no more bridging to configure for the main modes/slaves.)
  • The IDE hard disk drive uses the flat cable that has the width of around 18’’. But in the case of the Serial ATA a four wire cable is used that has a length of one meter. This is advantageous in comparison with the Parallel ATA (IDE) since the cables and the connectors that are used take less space than their counterparts in the Parallel ATA.
  • IDE platform is designed for normal / old computers, this way you transmit everything in parallel, and the IDE controller takes most of the work on transferring data.
    On Serial-ATA (SATA) the controller leaves most of the processing work to the CPU. (So if you have a small CPU you will prefer a IDE disk instead SATA, without mention that in old computers you will not have the SATA connector.)
  • IDE platform has it's top speed at 133 MB/s for ATA 133 (There is ATA 33, 66, 100 and 133) and for SATA you get 300 MB/s (SATA 1 have 150 MB/s, and SATA 300 MB/s).
  • Also, the SATA disk now are cheaper and bigger that any IDE disks.

Source: Various sites.
I have tried to be short & precise; have used simple language. Still queries n comments are welcome.

(PS: M blogging for the 1st time. please don't be too harsh!)

There are quite a few manufacturers that are producing different BIOS chips: Award, AMI, Phoenix, and IBM. Most commonly you can find AMIs, AWARDs and PHOENIXes. The BIOS is stored on a ROM chip.

These ROM BIOS chips can be of different measurements and look different from each other. Check out the following two types of chips. The one on the left is an AWARD (as stated on the sticker), while the one on the right is a Phoenix chip.