The Software and Hardware Components of the Watson Supercomputer - Case Study Example

IBM Super Computer, Watson Introduction IBM Watson is an artificially designed, intelligent system, which is d for the IBM founder Thomas J. Watson. Watson computing system was built by IBM team for meeting the grand challenges. During 2011, IBM Watson supercomputer competed and won against two competitors, namely, Ken Jennings and Brad Rutter in the challenges of Jeopardy. This supercomputer has the power to retrieve a humans ability through using natural language with a higher accurateness, speediness and confidence. It uses technic of natural language processing as well as artificial intelligence in order to find solutions of various sophisticated questions. Watson computer system will able to access structure and unstructured contents of millions of pages that are stored in disk. Watson through using the capacity of machine knowledge, power of statistical analysis and language processing recognizes and discovers the evidences of the problems. Moreover, it compares the most prominent possible solution, through positioning its confidence in its accuracy and responses in a very quick time (1IBM, n.d.). Correspondingly, the paper intends analyze both the software and hardware components of the Watson supercomputer. Additionally, the uses of Walton’s application in the other environment like health care and communication ecosystem will be evaluated. Discussion Processors IBM Watson is powered with 3000 processors, which enables the computing system to run algorithms depending on ‘Natural Language Processing’ (NLP). This processor on the other hand helps computer in translating the human language (high level language) in the form of the data that can be utilized as input in the computer system. This objective can be achieved with the help of the processor by reducing letters, words, or symbols into a gigantic database known as information. Additionally, through using 3000 processor enables IBM to deliver high quality, outstanding price-per-performance with processing data quickly and more efficiently. Server based on 3000 processors provides platforms, which can support lowering the workloads and total cost of ownership. It gives more reliability over the performance through focusing on business instead of computers (Schauer, 2008; Chapman, 2005). In this context through doubling, tripling or quadrupling the number of processors in the Watsons computing capabilities will significantly have the effect on the computer processing. With the increase in the number of processor counts, the execution time of the data bits will gradually decrease, thus improvising the overall performance of computing platform. The complication regarding parallel processing can be significantly minimized. In addition, the platform equipped with multiple-processors will be capable of handling large amount of data without increasing the overall load on the system (Schauer, 2008; Chapman, 2005). RAM The amount of availability of RAM has significant impacted the performance of the computer. With the 16 terabytes (TB) of RAM in Watson the system helps the system to increase the response time in the processing and make the system more productive. Furthermore, such a high amount of RAM enables the system to use the virtual memory up to an amount of reserved space for the overflow of RAM. Additionally, it also assists in delivering the overall operational efficiency through the large scale processing along with the memory-intensive applications. Thus, it can be concluded that along with the above benefits it empowers enterprises through consolidate applications in far more reliable, scalable and cost-effective system (ParetoLogic Inc., 2014; 6IBM, n.d.). Subsequently, adding extra RAM into the system will supports to increase the performance respectively. However, it significantly requires researching the pre-installed RAM, technicality, number of open slots in the system and the kind of RAM installed earlier. Adding extra RAM allow Personal Computer (PC) to cope up the problems and upgrade without any rebooting issue (ParetoLogic Inc., 2014; Monckton, 2013). Networking The communication network between the servers enables the database server and the various clients to transfer the necessary information among the others. This connection between the servers is known as a network. IBM communication network assists the system and enterprises having the windows environment. Based on this this communication network of the server’s customer and the business partner can highly establish the required client/server application of networking protocol with the high performance and high quality communication. Watson platform is made up 90 IBM Power 750 cluster servers, with additional I/O control nodes and network. The cluster network enables to perform work even if the one server is not working properly. Thus, the cluster communication network is highly beneficial and provides users to access the necessary services for the high time while on the other hand reduce unprepared outages. The network become highly reliable as it reduces the frequency of system failure. Additionally, through the least number of failure, the system will able to perform with high scalability along with higher transmission speed and capable of adding resources with far improve performance (Microsoft, 2014; 5IBM, n.d.). Data Storage The major advantage of Watson’s system is that it pre-stores massive amount of data in its memory. This data are stored using the software like Java, C++, Apache’s UIMA and Prolog. The system sources wide ranges of information including dictionaries, encyclopedias, thesauri, articles and other information’s that are installed by IBM. In the TV show Jeopardy! Challenge, Watson worked in an efficient and speedy manner regardless of being offline. To meet the necessary requirement, it was provided with millions of structured as well as unstructured content pages which consumed only 4 TB of overall disk storage this include ‘taxonomies, ontologies and databases’ (DATAVERSITY Education, LLC, 2013). IBM DeepQA Traditional QA system work through translating the natural language interfaces into the formal, structured query form and further issuing it against precompiled database in order to discover significant solution. Designing the philosophy of DeepQA meets the drawback of the traditional QA system albeit the system works on traditional manner to formally represent knowledge. Watson rather than finding the complete meaning of a question, it evaluates the answer of question through the minimum knowledge of the question. In this context, forming the DeepQA architecture in Watson significantly requires integration of many components, which faces substantial challenges, for instance the system needs to require regular monitoring and maintenance with high precision, accurateness, complex language and speed. The system also requires massive parallelism through taking into the consideration of multiple hypothesis and interpretation. At the same time, to the system requires many experts for its functioning such as for integration, application and evaluation. Furthermore, no component will decode answer on its own, so it requires pervasive confidence estimation to learn how to stack and chain the scores. Additionally, system will have to incorporate shallow and deep information’s to leverage ontologies (Ferrucci et al., 2010). SUSE Linux Enterprise Server 11 O / S IBM Watson utilizes the SUSE Linux Enterprise Server to work powerfully and provide proper business solutions. SUSE the pioneer in open source software help Watson to provide reliable, interoperable solutions that benefits the system to increase agility, reduce costs and manage complexity. SUSE Linux Enterprise Server 11 O/S is the latest Linux server operating system. SUSE Linux as the O/S provides the Watson to deliver advance features with superior performance along with maintaining the overall enterprise quality and compatibilities of the applications. Additionally, it enhances and most prominent stability with the highly secure foundation for the workloads built to power mission-critical workloads in physical, virtual and cloud environments. The main reason behind the use of SUSE in the Watson supercomputer is that it mitigates the pitfalls in the other operating systems and runs unreliable, cost-effective as well as secure manner. Furthermore, this OS runs five processor architectures in the real-time computing with great availability clustering (SUSE, 2014; 2IBM, n.d.). The SUSE Linux OS works in the differently with the other server OS in terms of its affordability, interoperability and vendor recommendations. The SUSE Linux OS doesn’t have any license cost as the other OS and can be downloaded using the SUSE home site. The cost with respect to maintenance and support is also quite low. The paid subscription of the server comes with the major benefits such as free upgrades, security fixes, patches and technical support. The server is uniquely designed in order to provide maximum interoperability with the environment of windows. On the other hand, Microsoft and SUSE since the year 2006 are working together in the field of ‘cross-platform virtualization’, ‘directory federation’ and system management to meet the requirements of customers more effectively. Furthermore, the Linux and the Microsoft provides environment enabled OS to communicate and share data in easier way. In this context, it can be affirmed that low-cost, associated with licenses requirement, maintenance and support along with the other hardware and system management tools makes SUSE different from other OSs (SUSE, 2014). Challenges of Data Accessibility in a Clustered Environment A Hadoop framework use of ‘Hadoop Distributed File System’ (HDFS), based on the Java file system which enable the entire system to store and access quickly, both structured and unstructured data across the large cluster server. Additionally, it doesn’t require any transformation or schema changes as the traditional database. Moreover, Hadoop is less expensive, easily scalability make it easier in the framework of analytical workloads on the real time basis. Apart from having the significant benefits of using the Hadoop framework, it may prove to be challenging in some respect. As with the changing demand of the computing and analytic company need to equally increase the requirement of storage repositories this large number of complex data that requires new technology to interpret. This interpretation requires constant analysis through close examination of the necessary facts and predicting the behavior which is quite complex task that may lead to reliability issues. Additionally, it involves running the huge, costly and complex systems in order to generate power. Each generators of the system holds difficult sensors that monitor frequency, current and other significant operating features. While operating a single generator requires careful attention towards streaming of all data, which may create the risk of data errors or failure in the entire system. The complexity and large amount of data availability has changes the enterprise thinking and in order to meet these requirements, new analytical capacity will be necessary. This will significantly necessitate regular upgrade, scalability, flexibility and cost based on the requirement of new architecture data warehouse. Moreover, with the constantly increasing availability and requirement of online data, the organizing the same will become quite difficult task for the system. The system must process those files and data with the exact attributes further this complexity fundamentally require high scalability platform for indexing those data (Cloudera, 2011). Apache UIMA Framework Unstructured Information Management Architecture (UIMA) is designed by the researcher of the IBM, which is a Java integration platform that helps to process the unstructured data analytically through using multi-modal software’s in order to form the structured data or knowledge. Conceptually, UIMA Java platform is characterized into the two major features i.e. declarative descriptors requiring metadata concerning on the each resources, like analysis engine resource, and common data structure, like the Common Annotation Structure (CAS). Furthermore, the UIMA Java is comprehended as ‘extensible’ and ‘scalable’ of UIMA framework, which enable the system to support applications in the various processing stages through acquiring the unstructured information in the form of the analysis and finally utilizes the end results. Furthermore, the UIMA Java framework helps the Watson as through the two basic mechanisms that include using path to store the various files into the local file system and use space of application class to locate resources (SYS-CON Media., 2009) Watson Application in the Health Care and Communications Ecosystems Growing advancement in technology, led to double the amount of healthcare information every year. Watson in the healthcare sector helps clinician to navigate the various complications of human language and enable them to analyze various data in a quick succession. Watson supercomputer have the advantages to enduring new research studies, which will help to deliver proper healthcare to the patients through published reports and articles. Furthermore, the capacity of holding millions of pages will reduce the complexity associated with the healthcare service like incomplete diagnosis. Watson provides opportunity to the new generation Cognitive Computing System (CCS) and healthcare to improve medication and practices. With the capacity of enduring vast amount of information through using NLP, evidence based learning and hypothesis generation. Watson supports healthcare professionals to provide proper care (Kohn, 2013; Rogers, & et. al., 2005). It allows organizations to earn potential profit and progress to large extent. Nevertheless, it also empowers ecosystem innovators to collaborate with the growing areas with high usability, cooperation with people fundamentally through improve access of information (3IBM, 2013). IBM Watson in todays era is scalable with expertise human knowledge in wider dimensions. Watson platform is utilized as the library of cognitive technology. Consequently, the scientists and researchers team of IBM and Watson group are continuously engaged in building more cognitive components in the system. Furthermore, Watson expands interpreted language according to the requirements of people, which significantly benefits various enterprises. Thus, the platform can be utilized in IT ecosystem as it will open up new global competency and help to conceive and design various projects (Olavsrud, 2014). Computer Systems Processing Considered to be equal or Equivalent to Humans’ Ability to Think Watson analyzes the underline questions and determines best possible answer, through advanced NLP, reasoning, machine technologies, information retrieval, knowledge answering. Watson utilizes the IBM DeepQA framework in order to generate the hypothesis, analysis, evidence gathering and scoring. Furthermore, it is able to load millions of documents, like taxonomies, texts, encyclopedias, plays religious novels along with the other reference materials that is used as the database. It understands the relative questions and evaluates the answers in the form relevant documents. Additionally, rather than relying on the single algorithm, it utilizes number of algorithms to find the exact core those answers, helpful evidences of the answers, and genuinely evaluate using complex NLP. Watson technology in the today’s scenario is providing the benefits to number of industries like healthcare, retailers and traveler among the others. (Markoff, 2011; 4IBM, 2009) Conclusion From the above discussion, it can be comprehended that IBM Watson supercomputer having the power to retrieved a humans ability through using natural language with a higher accurateness, speediness and confidence. The system is built with different technologies along with the other database to work in effective, reliable and economic manner. Watson is built with the hardware components including DeepQA, SUSE Linux OS, Hadoop framework and UIMA framework, which allow it to perform operations in superior way. IBM application establishes the system in such a way that it could be applied in any field, which requires high information with regular analysis. It is worth mentioning that the field of medical and communication have expressively benefited with Watson. Furthermore, it is anticipated that IT ecosystem will also be benefitted considerably with the use Watson in the years to come. References Cloudera. (2011). Ten common Hadoopable Problems real-world Hadoop use cases. Cloudera Whitepaper, 1-14. Chapman, M. T. (2005). The benefits of dual-core processors in high-performance computing. Supercharging Your High-Performance Computing, 1-18. DATAVERSITY Education, LLC. (2013). The big data bot: the future of Watson. Retrieved from http://www.dataversity.net/the-big-data-bot-the-future-of-watson/ Ferrucci, D., Brown, E., Chu-Carroll, J., Fan, J., Gondek, D., Kalyanpur, A. A., Lally, A., Murdock, J. W., Nyberg, E., Prager, J., Schlaefer, N., & Welty, C. (2010). Building Watson: An overview of the DeepQA project. Association for the Advancement of Artificial Intelligence, 59-79. 1IBM. (n.d.). What is Watson? Retrieved from http://www.ibm.com/smarterplanet/us/en/ibmwatson/what-is-watson.html 2IBM. (n.d.). IBM and SUSE. Retrieved from http://www-03.ibm.com/linux/suse.html 3IBM. (2013). Announcing a huge opportunity to put the power of Watson behind your apps!. Retrieved from https://ibm.com/https/www-304.ibm.com/partnerworld/wps/servlet/ContentHandler/isv_com_nws_exec_carter-watson-ecosystem 4IBM. (2009). A computer called Watson. Retrieved from https://ibm.com/https/www-http://www-03.ibm.com/ibm/history/ibm100/us/en/icons/watson/breakthroughs/ 5IBM. (n.d.). IBM cluster systems. Retrieved from http://www-03.ibm.com/systems/clusters/benefits.html 6IBM. (n.d.). Extraordinary scale, capacity and bandwidth for the most demanding enterprise applications. Retrieved from http://www-03.ibm.com/systems/power/hardware/795/features.html Kohn, M. (2013). Innovator chat: how Watson can transform healthcare. Retrieved from http://www.theatlantic.com/sponsored/ibm-watson/archive/2013/04/innovator-chat-how-watson-can-transform-healthcare/274584/ Markoff, J. (2011). Computer wins on ‘Jeopardy!’: Trivial, it’s not. Retrieved from http://www.nytimes.com/2011/02/17/science/17jeopardy-watson.html?pagewanted=all&_r=0 Microsoft. (2014). Evaluating the benefits of clustering. Retrieved from http://technet.microsoft.com/en-us/library/cc778629(v=ws.10).aspx Monckton, P. (2013). Will upgrading RAM make my PC or laptop faster? PC upgrade speed boost. Retrieved from http://www.pcadvisor.co.uk/how-to/pc-upgrades/3464779/should-i-boost-my-pcs-memory/ Olavsrud, T. (2014). IBM looks to outsmart big data competition with Watson. Retrieved from http://www.cio.com/article/2379621/enterprise-software/ibm-looks-to-outsmart-big-data-competition-with-watson.html ParetoLogic Inc., (2014). RAM: A key to PC performance. Retrieved from http://www.paretologic.com/resources/newsletter/pc_performance_tip_ram.aspx Rogers, M. L., Patterson, E., Chapman, R. & Render, M. (2005). Usability testing and the relation of clinical information systems to patient safety. Advances in Patient Safety, 2, 365-378. SUSE. (2014). SUSE Linux enterprise server. Retrieved from https://www.suse.com/products/server/frequently-asked-questions/ SYS-CON Media. (2009). Adopting OSGi in Java application frameworks: A case study defining reusable software system components. Retrieved from https://prod.cloudexpo2010east.sys-con.com/node/449864?page=0%2C2 Schauer, B. (2008). Multicore processors – A necessity. ProQuest, 1-14. Read More