Introduction
After Watson improved their processor to become the world fastest processor, working with Watson process has made computer more efficient. In terms of operations, this process has improved work execution from threefold to tenfold using DSP. This processor has come with great advantages to computer users with the synchronization counterpart well recognized. In order to realize such process to be effective, Watson came up with chronological process paradigms. Such program can be configured without having a central control unit. Since increasing their process, Watson computing company capability is more efficient. The efficiency of a computer is determined by the processing speed. Therefore, since the process speed of Watson is 3000 processors, this means that it is the most efficient computer that can be advised by use in the modern world. This computer is efficient because it has the capability of creating good planning and adding multiple pipeline network connectivity and high storage capacity. This includes large caches with the ability to execute more than one function simultaneously. This computer can execute more than 48 threads within a spark of a clock tick. Adding to this, Watson super computer has increase capability of producing accurate weather prediction, help in exploration of energy sources and developed cure for major diseases.
Doubling the number of processors has been effective since its invention by Watson Company. Since doubling its process, the computer has been more efficiency to perform many functions all at the same time. It started with doubling-up, then Watson tripled up and now they are quadrupling-up on their processors. This helps the computer to perform the basic function at a high speed (Bhanot 2004). Quadrupling-up helps the process to depend on multiple sources of memory in order to dangle on data while others functions are being processed. The first source of memory that is necessary for this processor to be efficient is processor register where information that is being process is stored. In quadrupling-up, it depends on how you use your computer because the processor can improve your speed by the use of auxiliary hardware. Doubling up the number of processor makes a super computer more reliable because it has the capability of perform more functions at ones without failure (Craft 2012).
The number of Random Access Memory RAM has direct link to the processing speed of a computer. While process data page, the number of RAM determine how faster information will be displayed on your monitor. Initially, the function of RAM is to hold and store random information. This is influenced by the number of RAM Watson computer is capable of holding. Therefore increasing the RAM to a larger RAM will be more efficient in processing data page with no major failure because the improved RAM will be able to host information that is more random while other functions are being performed. Doubling up RAM will give Watson computer more ability to perform more functions while the data page is still being process. However, doubling up RAM will help Watson super-computer avoid computer crashing down or slowdowns while operation (Craft 2012). This is because large RAM helps computer run more than one programs at a clock tick thus making it more efficient. In addition, doubling up Watson RAM will help escape low memory computer troubles. Larger RAM supports running of many programs thus making it able for this super computer to hold and perform function with enough memory for use.
Watson computer uses a huddle of servers that requires a connection network from the server. There are certain requirements that are important for there to be clear communication. In this program, we examine Cloud Networking-as-a-Service (CloudNaaS) that lengthens the mode of transmission from the cloud beyond the server. The construct is designed to accommodate for address, middle box and virtual network, which means the transmission speed, is designed to be linked from the server using cloud connections. The logical requirement for the network matrix, which takes two, input from the cloud server, which is new request from Watson and hosting. In this process, two protocols are involved. That is, TCP and UDP protocols. These two protocols are oriented connections. TCP is a software connection line that does not rely on IP address. A connection can be established between two different locations without using a cable. This can be made possible as long as both computers agree to the connection, and both computers have TCP resources to process the connectivity. The connectivity is made possible through an application.
TCP has several characteristics that ensure clear communication between the server and the host. One characteristic is reliability. Reliability implies that connectivity is through a network stream and a sequence that is generated by the server. This can only be interrupted network duplication or reorder (Bhanot 2004).
Watson super-computer has the capability of storing a large amount of data. This is because of its improved memory capacity. Watson computer has a RAM worth 16 terabyte. What this means it that Watson computer cannot face challenges of low memory. RAM helps computer process different function at the same time. For example, Watson computer can obtain data, while at the same time one is writing while also there is a program that is running. All this function can run swiftly without any major interruption because RAM has the capability of supporting these functions (Craft 2012).
In the DeepQA architecture, there are several challenges that are cause because of integrating more than three components in DeepQA architecture. First challenge is embarrassingly parallel. This can only occur when there is no connection between the parallel functions. Second challenge is software lockout where a program that is expected to perform in a particular task is lockout while performing other functions. Thirdly, parallel slowdown may occur while a connection is trying to be made causing a slowdown in process of function of a supercomputer. In addition, deadlock can occur while two or more functions are being process causing one program to wait for another to finish installing first (Craft 2012). Adding to this, livelock can also occur while two or more programs are being processed causing one to stop performing.
SUSE Linux Enterprise Server (SLES) is an operating system developed by SUSE. This operating system is Linux based. The design of this operating system is specific to server testing of desktop computers, to run on servers, workstations and mainframe computers. Brand new versions of this operating system are routinely released every 3-4 years. The minor versions usually referred to as service packs are released after every 18 months. The most current of this operating system version SLES 11 SP3 was released on July 1st 2013.
IBM made a choice of using SUSE Linux Enterprise Server because unlike other operating systems Linux-based operating system is very consistent, protected and scalable server operating system being specifically advanced to command mutually virtual and physical mission-critical workloads (Padua 2011).
Another feature that made IBM to choose this operating system server is its interoperability and affordability. This operating system is cost effective as it ensures deliverance of main business service while enabling security of computer networks thus making it is and inexpensive to administer heterogeneous IT possessions. Furthermore, SUSE Linux Enterprise server has comprehends modernization with vigorous industrial standardized servers that are unlock, green and easy. SUSE Linux also has the capability of combining database and software and hardware into one package making it able to use high performance computing clusters to arrive at solutions to problems with intensive data (Bhanot 2004). This operating system server is offered for 1-32 CPU sockets and particular server.
In addition to all this advantages of SUSE Linux it is mainly used by IBM Watson due to the fact that it is the fastest operating system available on IBM POWER in respect to current SPEC standards. The POWER 7 platform is understood for its capacity to meet high capacity, high functionality and next to linear scaling.
IBM Watson is advanced to match as much as possible the instructions of natural language dispensation capability of human brain. This intention of developing IBM Watson to perform these instructions calls for advanced memory management, unmatched functionality on system attached with multi-core processors, advanced I/O capabilities and multi-pathing, native POSIX Thread Library (NPTL) and new processor support (Craft 2012). SUSE Linux Enterprise servers only have these capabilities thus being highly favorable, otherwise if other operating systems missing one of the qualities is used then the supercomputer risks failing.
Apache Hadoop framework has an integrated Grid engine meaning that processing is not usually performed on a single rack with data. This puts into doubt the credibility of Hadoop making it less reliable as its runtime suggests so. ETL solution provided by Syncsort that prolongs the competence of Hadoop makes it more scalable, ease to utilize environment data integration and relatively affordable. IBM puts in offer WebSphere eXtreme Scale together with the scalability of Apache Hadoop distributed data capacity making it very available and transactional. This functions in that it facilitates suspension of function and capability of computer to store data for the sake of allowing quick function access. Errors and failure of the all system may also occur making it difficult to access information or data from the computer.
UIMA facilitates applications to be compressed into parts for instance language identification, sentence detection of boundaries, language entity detection and language specific segmentation. One of the above distinguished components puts into use interfaces specified by the UIMA framework providing self-explaining metadata through XML descriptor files (Padua 2011). The UIMA framework runs data flow between them and these defined components. It manages this by designing data flow between components by ensuring efficient mapping of languages. Components are given written in C++or in java to facilitate their management. It additionally wraps components into network services thus being able to scale large volumes doubling-up processing pipelines above grouped network nodes.
Watson has been in use in various fields concerning its efficiency and reliability in performing multiple tasks. It has been utilized in health care systems and environmental ecosystems. The iteration intervention has identified the best way of learning curve. The efficiency is evident since at earlier scenario it required five weeks to learn but now one only requires one hour to capture it. It has reduced the system into a server and the effectiveness is seen. The Watson system learns in multiple ways, it reads, fine tuning, answering questions and earns from its own mistakes. In the use of health care the Watson system has been trained to understand medical terminologies for example it knows that asthma is known by 16 different terms. The answers replied are evident basing on probabilities. In this instance if a doctor gives in a particular system the system, suggest the best diagnosis. This is based on the readings made from the many documents that pertains the symptoms. In the US, research shows that in every five individual of the person is given wrong diagnosis (Padua 2011). In terms of accuracy, the Watson system automatically updates the current diagnosing procedures as compared to the error doctor make due to lack of diagnosing procedures that are current. In terms of exchangeability of components that ensure the rapid running of the system’s processor.
Watson system can be applied in the IT ecosystem. This is possible because the software caused by the Watson system is compatible to be used in the information technology ecosystem. As much as it concerns giving out information, for example the Watson system responds to the commands similar to the actions done in IT.IBM is the maker of the various computers often used in the IT ecosystems. The drive of RAM, random access memory, which Watson uses a bigger one with 16 terabytes which makes it more powerful than any other computer. The use of networking system and the storage disk permits the IT to be applicable in Watson’s system. The system allows greater functioning o0f the application used (Padua 2011).
The Watson’s system supports and uses natural language and ability in performing tasks. It can answer questions promptly. The system is literally functioning like human beings since according to various feedbacks many corporate people have considered it to deal with customer service and take instructions given and programmed to do. Various business firms for example when a meeting conducted national venture capital association in San Francisco stated that plans were underway to build application software to boost the system (Padua 2011).
References:
Bhanot, G., Chen, D., Gara, A., Sexton, J., & Vranas, P. (2004). QCD on the BlueGene/L Supercomputer.
Craft, S. P., & Massachusetts Institute of Technology (2012). Stronger: The architects of a new intelligence.
Padua, D. A. (2011). Encyclopedia of parallel computing. New York: Springer.
