The Technical Analysis

Intel xeon 6 core server processor

2008-09-16T12:24:00.001-07:00

Intel unveiled Xeon 7400, the 6 core processor.Designed for visualization and server computing, Intel Xeon 7400 was developed with 16 MB L3 cache on monolithic architecture. That is all the cores are embedded on a single silicon chip called a die.

other features are

Socket compatibility with Intel 7300 series.
45nm Architecture
10% less power consumtion than its predecesor.
Virtualization
2.0 capability
Exclusively developed at Intels Banglore centre.

Chrome - Google Browser Review

2008-09-03T03:42:00.000-07:00

Chrome is the latest e-gadget from Google. We have to wait and see what so called 'revolution' it will make in the browser market. I downloaded Chrome from official Google Chrome website. It comes as a downloader of around 500KB which downloads rest of the browser from Internet.

Google had supported Firefox greatly till now. With the major market share being with Microsoft internet explorer and most techies opting for Mozilla Firefox available with infinite addons, its hard for Chrome to make a mark. But the name Google had always been linked with wonders. Don't make any predictions. The latest hit from Google was Google Docs. Will Chrome be the next Major hit? Lets wait and see...

Review Of Google Chrome

Do not seem that fast. Firefox is Faster, then Safari, chrome and IE. Might be because it is still a public Beta.
Nothing at bottom Part like a StatusBar. Just Kept Empty.
Plain and Simple Design.
"Resolving Proxy" would be the first term to get attention of anyone who uses Chrome. It is actually the Domain Lookup I think. It has to be faster.
Like Opera and Firefox Fastdial, thumbnail views of sites are available in each new Tab. Instead of manually setting the sites, it is automatically set as the most viewed sites. I love that change.
Flash sites are loading faster. I tried Nokia.com and miniclip.com
Even the main top bar or menu bar is not visible. Tabs has replaced it. What is Google gonna do without a menubar?
Design is too simple. Might make it hard to use.
A privacy mode browsing is possible without sending any of the information about the user to Internet. It will make context based advertising hard.
There are some improper loading on certain site. try http://www.nokia-caribbean.com/ at firefox and Internet explorer.

Click here to download Google Chrome

I'm not gonna change from Firefox to Google chrome even though I love Google products generally. These changes are not enough. Waiting for the next Major release.

MSI Wind review and specs

2008-08-31T09:38:00.001-07:00

MSI Wind is the new netbook or more precisely a subnotebook from Micro Star International. WIND is not just a brand name given to this 'ultra portable notebook'. It stands for Wi-Fi network Device. Wind is likely to cost at around 700$ in market. Both Windows XP and Linux versions will be available.

8 and 10 inch versions are available with a screen resolution of 1024 x 600. Unlike a Flash drive from segment competitors, MSI has included a 80GB Hard Disk with the Wind.
I test operated the white 1GB version for which I will certainly give a rating above Asus eePC in comparable range.

Other tech specs:

Chipset:
ICH7-M, Intel 945GMS
10"W (1024*600) LCD - LED backlit.
VGA:
UMA 950
1GB DDR2 Memory

Technical Analysis and review of MSI WIND.

Open office 3 review

2008-08-22T03:11:00.000-07:00

Open office 3 is striking hard at Microsoft office already. With the release of Open office.org version 3, the hit is being more severe. I just reviewed the windows version of open office 3.

The pack comes with 6 applications:

Writer - word processor
Calc - Spread sheet
Impress - Presentation maker
Base - Database application
Math - Equation Editor
Draw - Graphics program

It makes an almost perfect substitute for the costly Microsoft office. It has improved a lot this time. Start centre is a good addition which enables users to select the document they need to make from the welcome screen itself.

There is no other major changes to the total program. But there are numerous changes in each of individual applications. Overall changes are worth trying out.

Review Yourself and comment here

Intel Nehalem processor review

2008-08-21T02:23:00.001-07:00

Intel is all set to release their next generation processor Intel Nehalem boasting 8 cores. The speculations are so high that is is considered as "over" requirement for the current processor market. Nehalem is competing with the AMD Shanghai in the field. Both are set to be released around same date as per reports, this time no one is going to get the time advantage. ;-)

A Nehalem processor with 1GB DDR3 memory is said to posses same processing capabilityas a Intel penryn processor with 3GB DDR3. memory manufacturers are planning 3GB DDR3 for Nehalem based 32 bit OS and 12GB for 64bit. Nehalem samples ahve been showing good performances too, Intels claims seems right.

Major features of Nehalem summerised from Intels wesite are

Dynamic scalability
Design and performance scalability
Simultaneous multi-threading
Scalable shared memory

Nehalem , processor microarchitecture is was displayed in Computex 2008.

Intel atom processor performance review

2008-08-07T06:57:00.001-07:00

Intel atom is the new processor from Intel specially manufactured for mobile internet devices(MIDs). Previously codenamed Silverthorne and Diamondville, it is now given the new name atom. Major fields of application include Smartphones and ultramobile PCs which
are of relatively smaller size. The mobile technology which incorporates this processor is called Centrino Atom. Actually this is the first X86, a bit old I have to say, based processor after Intel Pentium.

At the cost of speed, Intel has made significant improvements in power consumption and cost. Average battery life is significantly extended by decreasing idle powen consumption. In short, Intel Atom processors are aimed at devices which need smallest power consumption as their primary feature. Power consumption problems had always been there with all intel mobile processors and this is not the first attempt to rectify it. But now Intel has significantly improved, if their technicalclaims are going to be true.

Having two versions, silverthorne will be called the Atom Z series and Diamondville will be called the Atom N series, according to Intel. Intel and Lenovo jointly announced new IdeaPad series notebooks based on Atom processor.
Intel atom is a 45 nm processor.
Intel is having a series of of competiters in the field. The major ones are Nvidia Tegra based line of processors for MIDs and the mobile platforms like Puma from AMD. Tegra is claimed to have more processing efficiency than Intel atom processor.
Intel has also announced the next generation atom processors based on Lincroft Architecture. It is codenamed Pineview. Intel will officially launch it in 2009.

Intel atom processor review

KDE 4.1 Desktop review

2008-08-04T13:24:00.001-07:00

This review focuses on KDE 4.1, the latest release the K Desktop Environment series. It was officially released on July 29th, 2008. There is no much hype regarding the new release since the older 4.0 was full of bugs. Still KDE remains one of the favorite desktop environment for GNU/Linux OS till date.

KDE 4.1 comes with significant positive changes in graphics and desktop experience. But still its not error free which is disappointing. The entire set of desktops, frames and panels are collectively called plasma in the new version(KDE4). This replaces the old Kdesktop in the older version. A new multimedia interface called Phonon has been added.This makes the KDE desktop enviornment independent of any particular media backend. Decibel is a new applicaition framework introduced in KDE to incorporate all the communication protocols in general to the K desktop. The amjor technologies introduced in KDE 4 could be summarised as follows:

Plasma : A desktop and panel widget engine
Phonon : A multimedia framework
Solid : A device integration framework
Sonnet : A spell checker
ThreadWeaver : A library to more effectively use multiprocessor systems

The release of KDE 4 and 4.1 invited mixed response from the web community. It was criticized on the grounds of instability of desktop and applications. But there are a lot of revolutionary changes made to KDE 4 for which it was praised by the KDE fans.

What are interrupts and interrupt handling?

2008-08-02T10:25:00.001-07:00

Interrupting is a mechanism by which the input or output devices of a computer informs the processor that it is ready to do its operation by raising a signal called interrupts.

It is highly related to the internal working of a computer. The input or output devices of a computer are connected to a processor by means of a bus. It is necessary to check the status of all input or output devices repeatedly to know if the device is ready to do its operation(read or write). This consumes lot of CPU time which can be used more effectively in doing other operations. Interrupts are used to avoid this.When an input device raises an interrupt signal processor comes to know that the device is ready. Then the processor will stop the current operation that is being executed by it and will begin to process the I/O device that caused interrupt signal. After knowing which device caused the interrupt, it will call a subroutine in response to it. This is called an interrupt service routine.

When an interrupt occurs, the current routine that was being executed by the processor, will be put into a stack in a known location. Later after servicing the interrupt the processor will come back to this routine. There can be interrupts in between interrupts. These are called nested interrupts. When a nested interrupt occurs, the last interrupt will be pushed to the stack again. Likewise, after servicing the last interrupt like that, the processor will 'run back' by popping previous routines that was saved in the stack. The process of saving instructions are done automatically by saving

When an interrupt occurs, the processor must inform the I/O device that it had identified the interrupt after doing so. This is done by an interrupt acknowledge signal that is sent back to it through the BUS.

Following is the typical scenario when an interrupt signal is raised:

The I/O device raises an interrupt request.
The processor stops(interrupts) the routine currently being executed.
Further interrupts are disabled.
The I/O device is informed that the interrupt is recognized. In response the I/O device deactivates the interrupt request signal.
The action required by the interrupt is performed by the interrupt service routine.
After completion, the interrupts are enabled and execution of the program that was previously interrupted was continued.

Vectored interrupts

Once a device raises an interrupt signal, the processor just knows that an interrupt is raised and not which device raised the interrupt. So it is necessary for processor to poll all the devices to know which device raised the interrupt. To avoid this, the interrupt signal itself can include the address of device raising the interrupt. Such interrupts are called vectored interrupts. The code supplied by the device will represent the starting address of the interrupt service routine for that device.

Google Knol - the unit of knowledge

2008-07-31T12:46:00.001-07:00

Google Knol was officially launched is out of its beta state. Knol - the word means a unit of knowledge. Google defines Knol as an authoritative article about a specific topic. All articles written in knol is verified by authorities in the topic before publishing. This makes Knol a trustable unit of Knowledge.

Knol is widely regarded as Google's attempt to beat Wikipedia. I am a huge fan of Wikipedia. But the fact that anyone can edit the contents of Wikipedia reduces its accuracy. It is said that there is a 1% factual error in every major article written in Wikipedia. This chance of error increases with less important topics. This is where Knol comes in use. Even though anyone can write a Knol, it is not published until it is verified by an authority in the topic.

There are a few hundreds of topics presently there at Knol. But being a Google product, it is very likely that it will keep increasing exponentially in future. None of the Google products had been a failure. Even though Google ads will be present in Knol, Google says that it will be minimum and content will be given sole importance.

Google Knol
You can Join Knol here or you can just Login with your Google account like any other Google services

What is Overclocking?

2008-07-29T03:06:00.001-07:00

The process of forcing any computer component like, processors, RAM, video cards and motherboard chipsets to run in a higher clock speed than the speed to which it was designed for by the manufacturer is called Overclocking. Overclocking is generally done by home PC users to boost the performance of their computer.

The main objective of overclocking is to improve the performance of low clock speed components to meet the minimum requirement of some new hardware or software. Some also overclock components to attain better performance from it. OEM systems generally disable this feature to prevent misconfiguration.

Even though it seems simple, overclocking is actually a very complex process involving high risk. One has to ensure sufficient voltage is given to the overclocked component. Excessive power can prove destructive too. Sufficient cooling measures should be taken as overclocking results in production of increased heat by the component.

News on overclocking

What is VoIP or BroadBand phone?

2008-07-24T13:27:00.001-07:00

VoIP or Voice over Internet protocol is a protocol that transmits voice over any packet swicthed network like Internet. The term VoIP is more generally used for actual transmission of voice through internet like broadband telephony, IP telephony, broadband phone or Internet Telephony. It actually refers to digital transmission of voice through Internet.

VoIP transmission can be done from pc to pc, pc to phone or phone to phone. It is largely dependents on the service provider who provides this broadband phone facility. To become the customer of IP phone, you need a computer with a broadband connection or a special type of phone with VoIP adapter.

The major advantage of VoIP is that calling rates for long distance calls are considerably reduced. VoIP calls are also generally more secure than traditional phone calls since the data is encrypted and transffered over network which is not easy to be tapped.

What does the project fork term mean in open source terminology ?

2008-07-23T11:58:00.001-07:00

When the source code of one open source software is copied entirely to modify and make a new one, it is called a fork. Usually forks are produced as a result of efforts to produce a new product with different point of view (goal) than the original product.

Forks are most common in Linux distributions. The most common example of fork is the Ubuntu project. It is a famous freely shipped Linux distro. But many of the non-techies dont know that it was forked from the first major Linux distribution in the world, Debian. Many of us even don't know that such a distribution exists. While in this case forking is friendly, many cases it can result in a war between forked and the original product.

Forking is one of the major drawbacks of the open source license. One can simply copy and change its name to make an entirely new product and can distribute. Such a situation is harmful for the software community since it duplicates the efforts put to develop the software and creates confusion among users to choose the product.

Fear of forking essay

What is MD5 Encryption?

2008-07-20T04:40:00.001-07:00

MD5 or message digest 5 is an encryption algorithm widely used for password comparisons. Basically MD5 is a hashing algorithm which takes a message of any length as input and produces a 128 bit or 32 hexadecimal digit digest as output. Even a very small variation in the inout string like one character can produce significant change in the output MD5 encrypted hash.

MD5 is a one way algorithm. That means a reverse algorithm of MD5 cannot decode the message digest back to its original message.

MD5 was designed in 1991 by Ron Rivest.

JPEG image compression technique

2008-07-19T11:20:00.001-07:00

I just completed reading the official FAQ of JPEG from faqs.org. JPEG is actually 'lossy' image standard widely used for compression of images. It stands for Joint Picture Experts Group which is the team who wrote this standard image file format from IETF.JPEG is mostly compared to GIF which has its own advantages and disadvantages over jpg.
The major problem associated introduction of with JPG was that the original group who worked behind the JPG format could not lay down a "standard" format due to certain internal limitations. Therefore major vendors manipulated the technologies behind the compression system to their own will even though basics remained the same.

JPEG uses the limitations of human eye such as inability to differentiate hue changes than changes in brightness. So a JPEG image scanned using a machine or software is likely to find errors but when viewed through human eye, it is almost unable to distinguish between the original image and compressed image. This make JPEG a lossy standard. But this small loss is compensated by the compression ratio. A JPG image is compressed to a ratio of 1:20 without visible changes in quality. If ready to give up a little details, one can compress this to 1:50 or even more. This makes JPG the favorite standard of web masters. Because during transmission over a network priority is transmission time rather than image quality. Since compression ratio is more, a JPG can be transmitted over internet with greater speed.
Unlike GIF, which holds 8bit data per pixel, JPEG holds 24bit. So that converting a true 24 bit data into a web standard image format is more viable with JPG. Experts say that GIF is very likely to be extinct in near future.

JPG compresses cartoon or black and white images poorly compared to GIF which is a limitation of JPG. Another problem arises when the sharp edges occurs. Dithering can take place at sharp edges in a jpg image.

JPG losses more data with repeated compression and decompression. So one should never try to compress a file that was once compressed.

There are two types of JPG mainly. They are baseline and progressive. In baseline all the image is read at once by the decoder. This make the image visible at once. But in a progressive one each image is split into various sections and are scanned differently. So for a viewer, it would seem that a progressive scanning results in faster loading but actually the time taken to completely view the file is same.

A JPEG file can not be made transparent at specific locations in a file unlike GIF. This is because of the fact that JPG is a lossy format.

IPv6- Internet protocol version 6

2008-07-19T11:17:00.001-07:00

IPv6 is an advanced implementation of internet protocol to include more space and thus more flexibility in assigning an IP address. Before learning about IPV6 I had doubted how how can one assign unique IP to each internet user in the cyberspace when the internet usage is going to explode in the years to come. And that was the same reason I Googled IPV6. And I'm sure my doubts are cleared now. It will be same for you if you know the number of combination of IP addresses possible with the next generation version of internet protocol name IPV6.

The exhaustion of IPv4 was predicted in 1992 and the call for a next generation internet protocol was initiated by IETF(Internet Engineering Task Force) in 1994. It was called IPng or Internet protocol next generation.

IPv6 supports 2^128 address since it uses a 128 bit code. In IPv4 it was 32 bit. An IP address in IPv6 is written in 8 groups containing 4 hexadecimal digits each.If any of the group if complete zeros it may be omited and can be replaced by '::'. Leading zeros can also be omitted as in IPv4.Thus a sample of v6 IP will be

2001:0db8:0000:0000:0000:0000:1428:57ab

In URLs version 6 IP should be enclosed in square brackets as in
https://[2001:db8::1428:57ab]:443/

Features of IPv6

The major improvement of Ipv6 over its predecessor was the larger address space possible.It completely avoids any kind of exhaustion of IPs and requirement for subnetting of the IP address.But this comes with a cost. That is this creates a bandwidth overhead.
IPv6 also supports a stateless address auto configuration. This means that whenever an IPv6 host is connected to a routed IPv6 network the IP address will get configured automatically. It also supports statefull address configuration as in IPv4 via DHCPv6.
IPv6 supports a static link-local address apart from the global address that applications generally use.
IPv6 can send and receive packets over the limits of up to 4GB while IPv4 had support only for 64KB.
Forwarding is made faster in Internet protocol version 6 by avoiding the checksum field implemented in IPv4. In IPv6, the link layer and transport protocols does the error checking thereby transportation speed is improved.

Primarily there are three types of IPv6.
namely,
Unicast Addresses
Multicast Addresses
Anycast Addresses

A Unicast address identifies a single network interface. Multicast identifies more than one. In a multicast address a packet sent to it is delivered to all interfaces Identified by it.Anycast addresses are assigned to different nodes for more than one interface.

Hashtable and collision resolution methods

2008-07-16T00:47:00.001-07:00

Hashing is basically a data searching method implemented using a data structure called hash table or hash map. A hash table includes a set of keys and values. If the key is given(for example an employee name in an employee database) one can easily find out the records associated with the key. In the hash table the keys are mapped to the location of values(data to be searched) or records using a hash function. That is the key is converted into memory address by passing it through a particular function called hash function. the records are stored in those locations indicated by memory addresses. later when a lookup or search is required, the given key is again passed through the hash function to obtain the location of records where it was previously stored. This eliminates the need for searching the entire records to choose the right one.

Collision :
A state of collision occurs when the hash function maps two different keys to same location. Thus the corresponding records can not be stored in the same location. The methods used to solve the problem of collision is called collision resolution. The two most popular method of resolving collision are,
Separate chaining
Open addressing
In open addressing new positions are computed using a probe sequence once the collision occurs and the next record is stored in that position accordingly. A link list is used to resolve collision in the second method called chaining.

What is Blade computing and Blade server?

2008-07-15T03:35:00.001-07:00

A Blade server is a server chassis containing server computers called server blades which are designed for high performance in less storage space. The difference between ordinary rack mount servers and blade servers is that all the power needs of the entire system is provided by the chassis as one single unit thereby eliminating the need for separate power cords for each server maintained. Other cable requirements are also eliminated. Thus a blade server saves space, which is a significant factor in modern computing era.

The consolidation of power supplies also reduces the total power consumption since individual power supplies to each server is not required. All the resources are centralized which results in lower management cost.It is easy to upgrade and add new servers. All that is needed is to slide the new servers into chassis space.

Each server blades contain 1 to 4 processors with ample memory racked in the blade chassis. One can also add additional memory modules outside the system. The entire system can be centralized by removing memory modules from server blades and adding one single external memory module. This can eliminate data redundancies too.

Server management tools and virtualization software comes in the software part which enables the server administration and ensures proper and maximum server utilisation.

Intel Centrino 2, next generation mobile technology

2008-07-12T13:38:00.001-07:00

The anticipated Intel Centrino 2 is finally on market. A project previously codenamed Montevina is now officially named Centrino 2 by intel. It is the first mobile platform on market based on Intel's latest Penryn CPUS. is, the Intel Centrino 2 is based on the advanced 45 nm technology which is a significant improvement from the old 65 nm chips.

The CPU speed will be ranging from 2.4Ghz to 2.8 Ghz.Other features are support for DDR3 ram , 802.11n and WiMAX. The Centrino 2 is said to be a big improvement while battery life and heat dissipation is concerned.It will run on 25 watt power rather than 35 of older ones.

What is SUN Spot?

2008-07-12T12:35:00.001-07:00

Sun SPOT (Small programmable object technology) is basically a Java programmable sensor or communication device developed by Sun Microsystems. Basically a small device of embedded programming which can be connected to a computer using a base station via USB. The SPOT contains a light sensor, temperature sensor and an accelerometer. These parameters can be programmed to produce a colour variations in a LED on the SPOT(a handheld device available with the kit).

The SPOT kit is basically the first step of an anticipated next generation technology by SUN for java based hardware systems. The most interesting part of this is that the entire system is opensource. It runs on a small version of Java virtual machine called Sqwak. We can use any IDE like netbeans to create applications for sun spot.

Technorati Tags: sun spot, sun micrsystems, sun, embedded systems,