Transcription
IBM SoftwareInformation ManagementData SheetIBM Netezza AnalyticsThe advanced analytics platform inside everyIBM Netezza applianceCustomers useIBM Netezza Analytics to: Predict with more accuracy Deliver predictions faster Respond rapidly to changesHighlights: Serious analytics – Answer questionsthat were previously too complex,required too much data or took toomuch time to analyze Data exploration – Discover subtlepatterns within ever-growing datavolumes to answer interrelated andcomplex business questions Analytics on-demand – Quicklyrespond to dynamic businessconditions to choose the bestcourse of action Simple to use – Build models andscore data using existing analyticstechnologies, including popularanalytic packages and languages High-performance – In-database,parallelized algorithms takeadvantage of IBM Netezza’sAsymmetric Massively ParallelProcessing (AMPP) architectureToday, enterprises are confronted with growing data volumes, morediverse data formats and sources, and increasing demands for speedyanswers to increasingly complex and interrelated business questions.Welcome to the world of “big data.” Advanced analytics helps businessesmake sense of this big data and get the answers they need to make betterdecisions and stay competitive. IBM Netezza Analytics is an embedded,purpose-built, advanced analytics platform — delivered with every IBMNetezza appliance — that empowers analytic enterprises to meet andexceed those business demands.IBM Netezza Analytics’ advanced technology fuses data warehousing andin-database analytics into a scalable, high-performance, massively paralleladvanced analytic platform that is designed to crunch through petascaledata volumes. This lets a large number of users ask questions of the datathat could not have been contemplated on other architectures. IBMNetezza Analytics is designed to quickly and effectively provide better andfaster answers to the most sophisticated business questions.IBM Netezza Analytics is IBM Netezza’s most powerful advancedanalytics platform that provides the technology infrastructure to supportenterprise deployment of in-database analytics. The analytics platformallows integration of its robust set of built-in analytics with leadinganalytic tools from such vendors as Revolution Analytics, SAS, IBMSPSS , Fuzzy Logix, and Zementis, on IBM Netezza’s core datawarehouse appliances. IBM Netezza pioneered the modern datawarehouse appliance and has customers worldwide that have realized thevalue of combining data warehousing and analytics into a single, highperformance integrated system. IBM Netezza Analytics enables analyticenterprises to realize significant business value from new business modelsand helps companies realize both top-line revenue growth and bottomline cost savings.
IBM SoftwareInformation ManagementData SheetIBM Netezza offers a distinctive and simple-to-use approachfor serious analytics. Traditionally, analytics are built anddeployed on separate analytic servers. This elongates theend-to-end time from model inception to deployment, andalso requires data movement, from a data warehouse or otherdata sources to the analytic server. In addition to this processtaking too much time, it is inefficient, limits the data used,constrains the scope of the analytic modeling, and impedesthe ability to experiment iteratively.Every IBM Netezza data warehouse is delivered with a libraryof pre-built, in-database analytic functions that can beaccessed through any SQL compliant interface or any of theother supported languages. Additionally, customers candevelop new capabilities using the platform’s user-definedextensions. It is an easy to use extensible platform thatsupports multiple tools, languages, and frameworks.With IBM Netezza Analytics, analytic models can be built anddeployed right where the data resides – in the data warehouse.By doing this, the time it takes to build and deploy analyticmodels throughout an enterprise can be significantly reduced.By shrinking the time from model inception to modeldeployment, companies can move to enterprise-wide, factbased decisions by infusing more of their decisions withinsightful on-demand analytics. IBM Netezza Analytics capabilities: Data exploration and discoveryData transformationModel buildingModel diagnosticsModel scoringThe IBM Netezza data warehouse appliance — a powerfulparallel computing platform — is fully exploited by IBMNetezza Analytics to deliver high-speed, scalable analyticsprocessing. The appliance uses the high-speed throughput ofthe Asymmetric Massively Parallel Processing (AMPP)architecture to maximize speed and efficiency for in-databaseanalytics processing. The AMPP architecture is a blade-basedIBM Netezza Analytics 2.0IBMInfoSphereStreamsTanay GPUApplianceby DevelopmentKitUser-DefinedExtensions(UDF, UDA,UDTF,UDAP)LanguageSupport(MapReduce,Java, Python,R, Lua, Perl,C, C ,Fortran,PMML)Apache Hadoop3rd PartyIn-DatabaseAnalyticsIBM athematicalSAS 9.3 IBM SPSSRevolutionAnalyticsGeospatialPredictiveDBLytix byFuzzy LogixUniversalPMML PlugIn by ZementisTime SeriesData MiningIBM Netezza AMPP PlatformFigure 1: IBM Netezza Analytics Architecture2EclipseStatisticsBI ToolsVisualizationTools
IBM SoftwareInformation ManagementData SheetNetezza’s simple appliance approach, all of an organization’sdata can be used to generate a finer set of results, helping todrive new revenue opportunities and gain a competitiveadvantage. By using advanced analytics on IBM Netezza datawarehouse appliances, the entire organization can realizevalue – from financial teams and lines-of-business, to sales andIT, to the executive office. This offers greater clarity for theentire business, and ensures everyone is leveraging the samedata, using all available data.DISKENCLOSURESSlice of User DataSwap and Mirror PartitionsHigh-Speed Data StreamingSMP HOSTSSQL, Query Plan,Optimize, AdminBy using IBM Netezza Analytics, organizations no longerhave to make a choice between large data volumes and seriousanalytics.SNIPPET BLADESData exploration(S-BLADES)Processor & Streaming DB LogicWith IBM Netezza Analytics, it is possible to move beyondtraditional business intelligence (BI) and ad-hoc reportingthat is based on historical data, and discover new ways tocreate value from data. Organizations now have the ability tobuild and deploy sophisticated analytic models that mirrorreal-world complexities more easily and effectively. They cancontinuously experiment, improve and tune analytic models todiscover trends and find ways to lower business risk, reducecost, increase revenues, and make fact-based decisions.High-performance databaseengine, complex analyticprocessing, streaming joins,aggregations, sorts, etc.streaming architecture that uses commodity blades andstorage, combined with IBM Netezza’s patented data filteringusing field programmable gate arrays (FPGAs), to deliverlarge data, high speed analytics. IBM Netezza hasconsolidated all analytics activity in a powerful and simpleappliance.Investigate new ways to create value from your data whiledoing concurrent, parallel model experimentation with IBMNetezza Analytics. Now, all data can be used instead of justsamples or aggregates, thereby improving accuracy andenabling more targeted decisions.IBM Netezza Analytics is purpose-built to simplify thebuilding and deploying of models for analytic enterprises thatdemand the highest performance on large, complex volumesof data.Telecommunications case studyIn the highly competitive world of telecommunications, the players needaccurate and up-to-date information to manage their businesses, exploitnew opportunities and keep customer churn to a minimum.Serious analyticsBusinesses are collecting and tracking information more thanever before, and are under increasing pressure to operatemore efficiently and effectively. The ability to analyze data,foresee outcomes and find ways to improve business is drivingcompanies to fully exploit advanced analytics. Making sense ofmassive volumes of data and turning it into meaningful resultscan be daunting or even technically unfeasible in companieswith traditional database technology. These systems are easilyoverextended just keeping up with the growth in user anddata volumes. The revenue assurance department uses analytics to identify revenueleakage and plug any gaps in the revenue chain. Product marketing and pricing teams can proactively plan and forecastthe effect of telephony tariff changes, and plan how to best react tocompetitors’ offerings. The credit services department uses analytics to spot high telephonyusage and proactively manage any potential credit problems beforethey arise.IBM Netezza data warehouse appliances crunch through large datavolumes and process in-database analytics to help telecommunicationscompanies be competitive and streamline processes.Analytics that once seemed impossible or impractical to runare now possible with IBM Netezza Analytics. With IBM3
IBM SoftwareInformation ManagementData SheetSimple to useHealthcare case studyThe IBM Netezza data warehouse appliance is easy-to-useand dramatically accelerates the entire analytic process. Theprogramming interfaces and parallelization options make itstraightforward to move a majority of analytics inside theappliance, regardless of whether they are being performedusing tools from such vendors as IBM SPSS, SAS, orRevolution Analytics, or written in languages such as Java,Lua, Perl, Python, R or Fortran. Additionally, IBM Netezzadata warehouse appliances are delivered with a built-in libraryof parallelized analytic functions, purpose-built for large datavolumes, to kick-start and accelerate any analytic applicationdevelopment and deployment.A healthcare provider was interested in predicting who was at risk fordiabetes. By taking a look beyond typical health parameters such asweight and family history, and adding more attributes to their modelssuch as financial background, the provider uncovered that a person’sfinancial situation does indeed impact their diabetes risk. By refining theiranalytic models, this healthcare provider was able to determine not only ifa person may develop diabetes at some point in his or her lifetime, butalso predict when the onset of the disease would take place (in one year,three years, etc).By identifying these trends, outreach and preventive care could beprovided to these patients at risk. This healthcare provider continues torefine and create models to uncover additional trends and improvepatient care leveraging IBM Netezza Analytics.The simplicity and ease of development is what truly sets IBMNetezza apart. It is the first appliance of its kind – packing thepower and scalability of hundreds of processing cores in anarchitecture ideally suited for parallel analytics. Instead of afragmented analytics infrastructure with multiple systemswhere data is replicated, IBM Netezza Analytics consolidatesall analytics activity in a powerful appliance. It is easy todeploy and requires minimal ongoing administration, for anoverall low total cost of ownership.Analytics on-demandOrganizations can get ahead of the competition by using IBMNetezza to predict, forecast and optimize business elements.Historically, the analytic process can be expensive andtime-consuming — it generally takes weeks to develop apredictive model from the data in a data warehouse. Once themodel is developed, it still takes hours – or even days in somecases — to execute on all the data, despite adding expensivehardware to the problem. The issue is further exacerbatedwith a growth in data volumes.Simplifying the process of exploring, calculating, modelingand scoring data are key drivers for successful adoption ofanalytics companywide. With IBM Netezza, business userscan run their own analytics in near real time, which helpsanalytics-backed, data-driven decisions to become pervasivethroughout an enterprise.IBM Netezza offers companies fast time-to-value forimportant predictive initiatives, resulting in a positive impacton the bottom line and in top-line growth. With IBMNetezza on board, organizations are armed with the mostaccurate intelligence to react more quickly and confidently toany opportunities or threats the market may present. Modelscan be quickly deployed, tweaked as needed, and dispensedconcurrently in multiples, while taking advantage of IBMNetezza’s parallel in-database technology. This enables thelargest data volumes to be handled quickly and efficiently.Retail case studyA global leader in behavior-based marketing solutions that providemanufacturers and retailers with the ability to execute targeted marketingprograms on the fly based on real-time market basket analysis andhistorical purchasing patterns, has seen great results from introducingIBM Netezza into its business. For example, in one campaign relying onPOS information to issue individualized coupons to customers, couponredemption increased by 30 percent using IBM Netezza.At a time when companies need to be as agile as possible toreact to changing market conditions and demands, aneasy-to-use system that runs blisteringly fast and analyzespetascale data makes a lot of sense.By leveraging the simplicity of IBM Netezza, this company has reducedthe number of DBAs required to maintain its data warehouseenvironment, significantly increased productivity, and IT analytics projectscomplete 5-10x faster.In addition, the company’s primary database storage space hasdecreased by almost 80TB since migrating to IBM Netezza, due to theelimination of all aggregate tables and indices. This reduction in storagealso decreased the corresponding data center footprint.4
IBM SoftwareInformation ManagementData SheetIBM Netezza data warehouse appliances use fieldprogrammable gate arrays (FPGAs), which have beenprogrammed by IBM Netezza to handle large volumes of datavery efficiently. These FPGAs filter out extraneous data as fastas it streams off the disk. This removes I/O bottlenecks andfrees up downstream components such as the CPU, memoryand network from processing unnecessary data, creating asignificant turbocharger effect on system performance.Digital media case studyA major digital media company, in the business of providing detailedanalytics to its clients, can provide right-time data, drive insightfuldecision-making, correctly measure marketplace dynamics andeffectively bridge the gap between retailers and manufacturers. Byleveraging IBM Netezza Analytics and IBM Netezza data warehouseappliances, the company can accommodate more customers now thanever before, and these customers can run custom-defined, ad-hocmarket analyses, whereas before they were limited to static views of themarket. Also, more data history can be retrieved and analyzed than everbefore and new data can be made available in near real-time. Analysesare unconstrained and have greater functionality and flexibility.The serious big math is performed in powerful multi-coreCPUs, where database primitives and complex analytics areexecuted on the filtered data stream. Analytic tasks are run asindependent processes operating on data streams across eachS-Blade. IBM Netezza’s parallel analytic engine harnesses thepower of all the computational cores in the appliance to offersignificant performance and scalability for serious analytics.IBM Netezza has helped reduce IT costs, thereby creating a moreeffective business model. With its advanced analytic capabilities, thecompany has a competitive advantage over its rivals.Focus on the business, not the process. Let your IBM Netezzadata warehouse appliance do the heavy lifting for you.Eliminate technology hurdles by leveraging the IBM Netezzaappliance to make your life simpler.High performanceWith IBM Netezza Analytics, you will have an appliance thatcan manage all of your analytic queries on massive datavolumes, all while taking advantage of the IBM Netezza datawarehouse appliance parallel processing platform for betterperformance. IBM Netezza provides you with the simpleappliance for serious analytics.IBM Netezza has created an extremely flexible analyticplatform that offers high performance at petascale. Bybringing analytics to the data, modelers and quantitativeteams can operate on the data directly inside the appliance,instead of having to move it to a different location and dealwith the associated data pre-processing and transformation.Analysts and modelers can take full advantage of the AMPParchitecture of IBM Netezza Analytics to ask the mostcomplex questions on all the enterprise data, without theinfrastructure getting in the way. Practitioners can iteratethrough different models more quickly to experiment and findthe best fit.Financial services case studyA financial institution needed to calculate value-at-risk for an equityoptions desk. The IBM Netezza platform was able to run a Monte Carlosimulation on 200,000 positions with 1,000 underlying stocks (2.5 billionsimulations) in less than an in-database analytics approach. This allowedthe financial institution to use the data where it resided as opposed tobuilding a parallel data-processing platform solely for performing theMonte Carlo simulation.Once the model is developed, it can be seamlessly executedagainst all the relevant data in the enterprise. The predictionand scoring can be done right where the data resides. Userscan get the results of prediction scores in near real time,helping operationalize advanced analytics and making itavailable throughout the enterprise.The combination of execution time and the elimination of latency requiredto move data between two platforms enabled the financial institution toinclude more variables in assessing the risk of investment strategies andto perform this assessment with greater frequency.5
IBM SoftwareInformation ManagementData SheetIBM Netezza Analytics PlatformIncluded in every IBM Netezza data warehouse appliancePre-built in-database analytics:Supports parallel model building and deployment/scoringTransformationsExecute data prep transformation in-database to gain significant performanceMathematicalPerform deep mathematical calculations in-database to leverage MPP processingStatisticsCalculate rich statistics without moving the dataTime seriesCreate forecasts and trends using deep, rich history to improve model accuracyData miningUse more, or all the data, to discover new and emerging insightsPredictivePredict with great accuracy and speed to move from batch processing to near real-time speed of thoughtanalyticsGeospatialImplement location-based analytics on big data with immediate feedbackSoftware Development Kit (SDK) includes:Multi-language supportDevelop with MapReduce, R, Java, Python, Lua, Perl, C, C , Fortran, PMML (export)Plug-in for EclipseBuild custom in-database analytics with easy-to-use, standard integrated development environmentUser-defined extensionsEmbed custom in-database analytics with user-defined functions (UDF), user-defined aggregates(UDA), user-defined table functions (UDTF), user-defined analytic processes (UDAP)Third-party applicationsAnalytic model development toolsIBM SPSS Modeler, Revolution Analytics R Enterprise for Netezza, Revolution Analytics R Visual ProductivityIn-database analyticsSAS, DB Lytix by Fuzzy Logix, IBM SPSS ModelerScoring enginesZementis Universal PMML Plug-In, SAS Scoring Engine for NetezzaConnectorsIBM InfoSphere BigInsights, IBM InfoSphere Streams, Cloudera, SAS, WPS Engine for NetezzaSpatial toolsESRI, Safe Software, Pitney Bowes MapInfo, Google Earth, Microsoft VirtualEarth, BIS2, IBM DataStage ,Informatica, IBM Cognos , Microstrategy, Business Objects, IntegoGPU acceleratorsTanay by Fuzzy LogixData integrationAb Initio, BusinessObjects/SAP, Composite Software, DataFlux – a SAS company, Expressor Software,Informatica, IBM Information Server, Oracle GoldenGate software, Oracle Sunopsis, WisdomForceData analysisIBM SPSS, Revol
IBM Netezza Analytics’ advanced technology fuses data warehousing and in-database analytics into a scalable, high-performance, massively parallel advanced analytic platform that is designed to crunch through petascale data volumes. Th
