Platforms A SearchDataManagement.com Directory

2009 EDITION

This product brought to you by: Welcome!

INTRO

SELECTING THE RIGHT DATA WAREHOUSE PLATFORM FOR YOUR ORGANIZATION

INDEX Welcome to SearchDataManagement.com’s Data Warehouse Platforms Directory. This directory is designed to be a valuable resource for those getting started with research or evaluating optimized data warehouses. Inside, you’ll find basic infor - 1010DATA mation about vendors and the platforms they sell. Each listing is accompanied by ASTER DATA a short description and a long description, including limited information about SYSTEMS functionality and product use. You’ll find products for businesses of all sizes as IBM well as products that can be deployed on-demand and on-premise. Use this list ILLUMINATE to get started with the evaluation process. For more information about any of the SOLUTIONS products or to speak to a sales representative, please visit the vendor website or product website. INFORMATICA

KALIDO SearchDataManagement.com will launch a series of directories throughout the

KOGNITIO year to address unique segments of the data management market. Want to see your product listed in one of our directories? Go here to submit a product . Need to CORP. update product or pricing information? Email us here . For questions for the editors or to make suggestions for improving the directory, write to us at editor@search - ORACLE datamanagement.com . CORPORATION

PARACCEL PERVASIVE Happy shopping! SAND TECHNOLOGY

SUN MICROSYSTEMS

SYBASE

TERADATA

VERTICA

ABOUT SEARCHDATA- MANAGEMENT .COM

METHODOLOGY

DATA WAREHOUSE PLATFORM PRODUCT DIRECTORY 2 Selecting the right data warehouse platform for your organization INTRO

SELECTING THE RIGHT DATA WAREHOUSE PLATFORM FOR YOUR ORGANIZATION

INDEX s companies consider I Quality data that is available. their ability to manage I Architectures that provide low 1010DATA the asset that is infor - long-term total cost of ownership. ASTER DATA I SYSTEMS mation, the data Good query performance that warehouse platform (and its results in increased interactive IBM Amanagement system, DBMS) is the usage. ILLUMINATE I SOLUTIONS most important decision point. The The ability to get to real-time feeds. platform is the foundational component I A platform to support mixed, GREENPLUM of the tool selections, the consultancy unpredictable workloads. INFORMATICA hires, the architecture, etc. In short, it I A scalable path forward as data KALIDO defines your information culture. needs grow. KOGNITIO However, in selecting the platform to

MICROSOFT CORP. support the data warehouse, organiza - tions are faced with an exponentially CRITERIA FOR DATA NETEZZA higher number of variations and distinct WAREHOUSE PLATFORM SELECTION departures from the traditional online The decision process should go well transactional processing (OLTP) data - beyond the usual consideration of the PARACCEL base management systems than ever operational DBMS vendor. Today, that PERVASIVE before. decision can be nuanced along several SAND TECHNOLOGY Over time, data warehouse data vol - potential requirements, including: umes will continue to soar as history I SYBASE accumulates, syndicated data is collect - Active loading of data and immedi - ed and new sources with more detailed ate access to loaded data. data are added. Furthermore, the com - I A mixed workload upon the data. munity consuming the data continues I The immediate availability of infor - to grow, expanding well beyond compa - mation. ABOUT ny boundaries to customers, supply I Cross-functional complexity. SEARCHDATA- I MANAGEMENT chain partners and even the Internet. The level of query concurrency. .COM Companies need to make sure they I The scalability needs of the plat - METHODOLOGY choose a proven platform not just for form. the initial, known requirements but also I The functionality of the DBMS. with the ability to scale to future, to-be- determined requirements. Given the state of the marketplace, Data warehouses, to be successful, the technical architecture for a data need to provide: warehouse platform should be:

DATA WAREHOUSE PLATFORM PRODUCT DIRECTORY 3 Scalable: The data warehouse plat - Available: The platform should sup - form should be scalable in both per - port mission-critical business applica - formance capacity and incremental tions with minimal downtime. Check on

INTRO data volume growth. Make sure the pro - “hot pluggable” components, understand posed platform scales in a near-linear system downtime requirements and SELECTING THE RIGHT DATA fashion and behaves consistently with any issues that might deny or degrade WAREHOUSE PLATFORM growth in all of database size, number service to end users. These can include FOR YOUR of concurrent users, and complexity of batch load times, software/ hardware ORGANIZATION queries. Understand the additional upgrades, severe system performance INDEX hardware and software required for issues, and system maintenance outages. each of the incremental uses.

1010DATA Interoperable: The platform should

ASTER DATA Powerful: The data warehouse plat - allow for integrated access to the Web, SYSTEMS form should be designed for complex internal networks and corporate main -

IBM decision support activity in a multi-user frames.

ILLUMINATE mixed workload environment. Check on SOLUTIONS the maturity of the optimizer for sup - Affordable: The proposed technology

GREENPLUM porting every type of query with good (hardware, software, services, required

INFORMATICA performance and to determine the best customer support) should allow for low execution plan based on changing data total cost of ownership (TCO) over a KALIDO demographics. Check on conditional multi-year period. KOGNITIO parallelism and what the causes are of MICROSOFT CORP. variations in the parallelism deployed. Flexible: The platform should be able

NETEZZA Check on dynamic and controllable pri - to provide optimal performance across

ORACLE oritization of resources for queries. the full range of normalized, star and CORPORATION hybrid data schemas with large num -

PARACCEL Manageable: The platform should be bers of tables. Look for a proven ability

PERVASIVE manageable with minimal support tasks to support multiple applications from requiring DBA/System Administrator different business units, leveraging data SAND TECHNOLOGY intervention. It should provide a single that is integrated across business func - SUN MICROSYSTEMS point of control to simplify system tions and subject areas. SYBASE administration. You should be able to

TERADATA create and implement new tables and Robust in-database management

VERTICA indexes at will. systems, features and functions: The platform should include DBA productiv - Extensible: The data warehouse plat - ity tools, monitoring features, parallel ABOUT SEARCHDATA- form should provide flexible database utilities, robust query optimizer, locking MANAGEMENT design and system architecture that schemes, security methodology, intra- .COM keeps pace with evolving business re- query parallel implementation for all METHODOLOGY quirements and leverages existing in- possible access paths, chargeback and vestment in hardware and applications. accounting features, and remote main - Ask such questions as: What is required tenance capabilities. to add and delete columns? What is the impact of repartitioning tables? Referenceable: There may not be one

DATA WAREHOUSE PLATFORM PRODUCT DIRECTORY 4 single reference that matches your ORIENTATION environment exactly, but you should be A row-oriented database is an imple - able to see a consistent trend across a mentation of relational theory. Data

INTRO wide range of references that enforces is stored as bytes, with all columns for what you are looking for in a data ware - a row stored in order. These bytes are SELECTING THE RIGHT DATA house platform. grouped by the several thousand—from WAREHOUSE PLATFORM Organizations should also consider 4,000 to 64,000—into data blocks. FOR YOUR vendor viability, especially in these This is the unit of input/output, an ORGANIZATION days of vendor consolidation. The ven - exception being the “pre-fetch” capa- INDEX dor’s financial stability, the importance of data warehousing to their overall 1010DATA business strategy and their continued “Organizations should ASTER DATA research and development in these SYSTEMS areas toward a well-developed and also consider vendor IBM relevant vision are all key components viability, especially in ILLUMINATE of a vendor’s viability in this critical SOLUTIONS decision. these days of vendor GREENPLUM consolidation.” INFORMATICA DATA WAREHOUSE —WILLIAM MCKNIGHT KALIDO ARCHITECTURAL CONSIDERATIONS KOGNITIO To enable the above criteria, companies MICROSOFT CORP. have some new options in the types of bility of row-oriented database man -

NETEZZA data warehouses available, many of agement systems (DBMS) sensing a

ORACLE which have been created in recent years pattern in the reads. A columnar DBMS CORPORATION owing to the high cost of ownership of is an implementation of the relational

PARACCEL the traditional approaches. theory, but with a twist. The data stor -

PERVASIVE Today, typical data management age layer contains a grouping of environments have several restrictions columns. For example, all the column 1 SAND TECHNOLOGY they’ve "learned to live with” in seeking values are physically together, followed SUN MICROSYSTEMS acceptable analytical performance. by all the column 2 values, etc. The data SYBASE Indexes, summary tables, cubes, utility is stored in record order, so the 100th

TERADATA executions like reorganizations and vari - entry for column 1 and the 100th entry

VERTICA ous de-normalizations are often intro - for column 2 belong to the same input duced into the environment. There's a record. point at which overcoming these chal - In a columnar database, all the same ABOUT SEARCHDATA- lenges ceases to be "easy" and afford - data—your data—is there. It’s just MANAGEMENT able and therefore ceases to get done. organized differently (automatically, by .COM But perhaps more important than these the DBMS). The main benefits include METHODOLOGY very real restrictions are the restrictions the ability to highly compress the data. in possibilities thinking and information A byproduct of the compression is the exploitation. The higher the untuned ability to perform columnar opera - performance of the environment is, the tions—like MIN, MAX, SUM, COUNT better. and AVG—very rapidly.

DATA WAREHOUSE PLATFORM PRODUCT DIRECTORY 5 DATA WAREHOUSE APPLIANCES MODEL tures an exponential uptake from SMP Data warehouse appliances have and clusters in the number of proces - emerged as viable short-list solutions sors and the sophistication of the inter - for new or refurbishing data manage - INTRO connect. MPP is the generally acknowl - ment efforts. Vendors now dot the land - edged ideal parallel architecture for SELECTING THE RIGHT DATA scape (with representation in analyst analytic query, and WAREHOUSE "quadrants" and "spectrums" and hun - PLATFORM data warehousing. FOR YOUR dreds of millions in venture backing) ORGANIZATION and they are worthy of any data ware - INDEX house professional’s understanding, “A true ‘data attention and consideration. warehouse in a 1010DATA The data warehouse appliance is a hardware/software/OS/DBMS/storage ASTER DATA box’ has remained SYSTEMS preconfiguration for data management somewhat elusive...” IBM requirements. Many utilize commodity components and some use open source —WILLIAM MCKNIGHT ILLUMINATE SOLUTIONS DBMS for a low overall TCO. These

GREENPLUM open source DBMSs provide a starting point for basic database functionality, PACKAGED ANALYTICS INFORMATICA and the appliance vendors focus on Finally, today, most large enterprise KALIDO necessary functionality enhancements. software vendors offer some type of KOGNITIO Performance of queries, especially pre-built reports or dashboards cen - MICROSOFT CORP. against large volumes of data, is distinc - tered on typical business functions such

NETEZZA tively impressive thanks to the auto - as Finance, Sales and Marketing, and matic parallelism many, but not all, ORACLE Supply Chain Management. In addition, CORPORATION appliances provide. Low TCO for a a number of smaller companies have

PARACCEL mixed-workload data warehouse envi - developed packaged analytics built on ronment is consequential with appli - PERVASIVE licensed platforms and targeted to spe - ances as well. cific industries, applications or special - SAND TECHNOLOGY ized functions, or to supplement other SUN MICROSYSTEMS third-party services. SYBASE PARALLELISM ARCHITECTURE While a true “data warehouse in a

TERADATA Parallelism is as important as columnar box” has remained somewhat elusive, orientation to query performance suc - VERTICA analytic packages bridge the gap cess, and massively parallel processing between off-the-shelf canned reports (MPP) systems provide the most paral - and a ground-up custom BI software. ABOUT lelism possible. MPP systems evolved SEARCHDATA- Such packages simplify the develop - MANAGEMENT from symmetric multiprocessing (SMP) ment process and deliver functionality .COM systems, which, in turn, evolved from with more cost certainty. Organizations METHODOLOGY uniprocessor systems. must complete a careful evaluation Sharing disk and/or memory across process to determine whether a pack - nodes creates overhead. Sharing noth - age is right for them, and they must ing minimizes disk access bottlenecks understand the various costs and bene - and is preferred in MPP, which also fea - fits of different options. With all of this

DATA WAREHOUSE PLATFORM PRODUCT DIRECTORY 6 information at hand, decision makers bilities, with emphasis on concur - will be better positioned to make the rency. best choice for their organization. 4. Compile qualitative information on

INTRO each offering, using published information, consulting, and refer - SELECTING THE RIGHT DATA DATA WAREHOUSE SELECTION PROCESS ences. WAREHOUSE PLATFORM Once the possibilities have been nar - 5. Weigh quantitative and qualita- FOR YOUR rowed down according to the evaluation tive recommendations on each ORGANIZATION of the above, a proof of concept should vendor. INDEX be done with your data. Data ware - house vendors all have the ability to Make sure each important criterion

1010DATA take in your data for an offsite proof of is covered in the POC.

ASTER DATA concept. These offsite data centers are SYSTEMS well worth using. You can “visit your

IBM data” in person in the process, if you CONCLUSION

ILLUMINATE wish. These days, the best answer to your SOLUTIONS You need to determine, and gather, data warehouse platform selection

GREENPLUM the representative data as well as deter - needs may not be as evident as it was in

INFORMATICA mine the representative data access the past. Given the high relevance of paths that you wish to explore in a proof data and information, investment and KALIDO of concept (POC). The POC should innovation continue in this area and KOGNITIO comprise these phases: there are a surprisingly large number of MICROSOFT CORP. options. Many platforms may work, but

NETEZZA 1. Prepare and gather data. at high short- and long-term prices.

ORACLE 2. Set up environment(s) at off-site Making a selection consistent with your CORPORATION location(s). goals as an organization and with a

PARACCEL 3. Validate data access tool connec - broad and detailed view of the land - I PERVASIVE tivity, performance, and load capa - scape is paramount to success.

SAND TECHNOLOGY

SUN MICROSYSTEMS

SYBASE

TERADATA

VERTICA

ABOUT SEARCHDATA- ABOUT THE AUTHOR MANAGEMENT .COM WilliamMcKnight , President, McKnight Consulting Group William functions as strategist, lead enterprise information architect and program manager for complex, METHODOLOGY high-volume, full lifecycle implementations worldwide—utilizing the disciplines of data warehousing, master data management (MDM), business intelligence (BI), data quality and operational BI. McKnight is a Southwest Entrepreneur of the Year finalist and a frequent best-practices judge; he has authored more than 150 articles and white papers and given more than 150 international keynotes and public seminars. His team’s implementations from both IT and consultant positions have won Best Practices awards. He is a former IT VP of a Fortune company, a former engineer of DB2 at IBM, and holds an MBA. He can be reached at [email protected] .

DATA WAREHOUSE PLATFORM PRODUCT DIRECTORY 7 Index at a Glance Click on the product name at left to jump to a longer description.

INTRO SAAS OR ON PAGE VENDOR PRODUCTS SERVICES PREMISE SW SELECTING THE RIGHT DATA WAREHOUSE 10 1010data 1010data 1 s PLATFORM FOR YOUR 10 Aster Data Systems Aster nCluster 1 s ORGANIZATION 11 IBM InfoSphere Warehouse s INDEX 11 illuminate Solutions iLuminate 1 s

1010DATA 12 Greenplum Greenplum Database 3.3 1 s

ASTER DATA 12 Informatica PowerCenter s SYSTEMS

IBM 13 Kalido Kalido Dynamic Information Warehouse (DIW) s ILLUMINATE SOLUTIONS 13 Kognitio WX2 analytical database 1 s GREENPLUM 14 Microsoft Corp. SQL Server 2008 s INFORMATICA 14 Netezza AppNexus s KALIDO 15 Netezza Data Warehouse Appliances s KOGNITIO

MICROSOFT CORP. 15 Oracle Corporation HP Oracle Database Machine s

NETEZZA 16 ParAccel The ParAccel Analytic Database (PADB) 1 s

ORACLE 16 Pervasive Pervasive Data Warehouse s CORPORATION

PARACCEL 17 SAND Technology SAND/DNA Access 1 s

PERVASIVE 17 Sun Microsystems MySQL 1 s

SAND TECHNOLOGY 18 Sybase Sybase IQ 15 s SUN MICROSYSTEMS 18 Teradata The Teradata Purpose-Built Platform Family s SYBASE 19 Vertica Vertica Analytical Database 3.0 1 s TERADATA 19 Vertica Vertica Analytic Database for the Cloud 1 VERTICA 20 Vertica Vertica Virtualized Analytical Database 1

ABOUT SEARCHDATA- MANAGEMENT .COM

METHODOLOGY

Vendor: Vendor/developer of product at directory press time; 1 SaaS or services indicates technology available as SaaS, hosted, on-demand, ASP and Web services; s On-premise indicates software or systems on premise; 2 Descriptions were written by the SearchDataManagement.com editorial team based on information gathered from vendor websites.

DATA WAREHOUSE PLATFORM PRODUCT DIRECTORY 8 1010DATA ASTER DATA SYSTEMS

1010data Aster nCluster INTRO

SELECTING THE 1010data’s data warehousing service Aster nCluster is built as a high- RIGHT DATA provides a high-performance alternative performance analytic SQL database WAREHOUSE PLATFORM to DW and conventional OLAP tech - for frontline data warehousing and FOR YOUR ORGANIZATION nologies, empowering organizations analytics. 1 s with a fast, reliable, cost-effective and INDEX easy-to-use capability to get the most COMPANY WEBSITE: www.asterdata.com out of data. 1 s FOUNDED: 2005 1010DATA SUMMARY: Aster nCluster is a high- ASTER DATA COMPANY WEBSITE: www.1010data.com performance analytic SQL database SYSTEMS FOUNDED : 2000 for frontline data warehousing. It’s IBM SUMMARY : 1010data’s service provides a designed to provide an always-on, ILLUMINATE unique platform that combines front- always-parallel MPP architecture with SOLUTIONS end usability with back-end functionali - the first-ever In-Database MapReduce GREENPLUM ty. Delivered as a managed solution, programming framework. “Always par - INFORMATICA 1010data allows analysts to quickly allel” means that Aster can deliver max -

KALIDO construct complex, sophisticated imum system performance and efficien - queries on very large datasets and get cy through network-optimized MPP KOGNITIO results in seconds, allowing organiza - architecture. “Always on” means that MICROSOFT CORP. tions to make quick, critical decisions. Aster can provide hands-free 24/7 NETEZZA 1010data’s service model provides a availability for data-driven applications. ORACLE low-risk, fully managed and supported And In-Database MapReduce is CORPORATION database solution that does not require designed to provide in-database analyt - PARACCEL additional personnel, has a short time ics done right—the analytic expressive - PERVASIVE to delivery and is relatively inexpensive. ness and flexibility of MapReduce com -

SAND TECHNOLOGY Since it is so quick and easy to inte - bined with the rich functionality and grate, enterprises experience high support of SQL. SUN MICROSYSTEMS speed-to-value and very rapid ROI . SYBASE PRICING: Aster prices are based on TERADATA PRICING: 1010data’s business intelligence amount of user data. Pricing starts at VERTICA service starts at $50,000; price is $100,000. Depending on the amount based on amount of data and the num - of data being analyzed, pricing can go ber and type of users to ensure proper from low six-figures to the seven-figure ABOUT SEARCHDATA- service levels. range. MANAGEMENT .COM

METHODOLOGY

DATA WAREHOUSE PLATFORM PRODUCT DIRECTORY 9 IBM ILLUMINATE SOLUTIONS

InfoSphere Warehouse iLuminate INTRO

SELECTING THE InfoSphere Warehouse is designed to iLuminate is a correlation database RIGHT DATA be a comprehensive data warehouse engine that is designed to contain WAREHOUSE PLATFORM tool, providing all of the capabilities all the data management tools needed FOR YOUR ORGANIZATION necessary to glean maximum return to build and manage an enterprise data from your information. s warehouse or . 1 s INDEX

COMPANY WEBSITE: www..com COMPANY WEBSITE: www.illuminateinc.com 1010DATA FOUNDED: 1896 (as the Tabulating FOUNDED: 2005 ASTER DATA Machine Company) SUMMARY: The iLuminate correlation SYSTEMS SUMMARY: IBM InfoSphere Warehouse is database engine uses value-based stor - IBM desinged to help organizations take the age (VBS), which stores data values ILLUMINATE next step in innovating with informa - only once and uses a sophisticated SOLUTIONS tion. It provides a powerful tool set indexing system to logically connect all GREENPLUM delivering access to structured and data correlations and relationships. The INFORMATICA unstructured information and delivering engine contains the tools needed to

KALIDO operational insights in real-time. InfoS - build and manage an enterprise data phere is designed to reduce costs by warehouse or data mart with automat - KOGNITIO freeing storage through a broader use of ed database design and no need for MICROSOFT CORP. compression and by making workload users to define requirements up front. NETEZZA management a standard feature of The database—actually the indexing ORACLE Departmental and Enterprise editions. system itself—is created on the fly dur - CORPORATION It’s also designed to optimize the use of ing the loading process and is inherently PARACCEL industry-standard XML-formatted data compact and fast. During the loading PERVASIVE by directly storing standard XML data process, the data itself creates its own

SAND TECHNOLOGY into the data warehouse, compressing data-generated schema and all the data XML data to save on storage costs and values are 100% indexed. This provides SUN MICROSYSTEMS improving performance when managing fast train-of-thought analysis. SYBASE XML records . 2 TERADATA PRICING: Price is based on the number of VERTICA PRICING: Vendor declined to provide records loaded; starts at $41,900 for up pricing. to 35 million records. Concurrent con - nection fees apply and start at $2,750 ABOUT SEARCHDATA- for up to four connections. MANAGEMENT .COM

METHODOLOGY

DATA WAREHOUSE PLATFORM PRODUCT DIRECTORY 10 GREENPLUM INFORMATICA

Greenplum Database 3.3 PowerCenter INTRO

SELECTING THE Greenplum Database is software built Informatica PowerCenter is a single, RIGHT DATA to support the next generation of data unified enterprise data integration WAREHOUSE PLATFORM warehousing and large-scale analytics platform for accessing, discovering FOR YOUR ORGANIZATION processing. 1 s and integrating data from virtually any business system, in any format, INDEX COMPANY WEBSITE: www.greenplum.com/ and delivering that data throughout FOUNDED: 2003 the enterprise at any speed. s 1010DATA SUMMARY: Greenplum Database utilizes a ASTER DATA shared-nothing massively parallel pro - COMPANY WEBSITE: www.informatica.com SYSTEMS cessing architecture optimized for busi - FOUNDED: 1993 IBM ness intelligence and analytical pro - SUMMARY: Highly available, high-perform - ILLUMINATE cessing. It allows a cluster of servers to ing and fully scalable, PowerCenter SOLUTIONS operate as a single database supercom - serves as the foundation for all data GREENPLUM puter—automatically partitioning data integration projects and enterprise inte - INFORMATICA and parallelizing queries—to achieve gration initiatives across the enterprise,

KALIDO very fast performance times. Green - including B2B data exchange, data gov - plum Database moves processing ernance, data migration, data synchro - KOGNITIO power as close as possible to the data, nization and replication, enterprise data MICROSOFT CORP. so processing always occurs in parallel, warehousing, integration competency NETEZZA delivering unmatched query and load centers (ICCs), master data manage - ORACLE performance. With Greenplum Data - ment (MDM), and service-oriented CORPORATION base, users have the power to answer architectures (SOAs). PowerCenter is PARACCEL complex questions, running analyses designed to deliver trusted, timely data PERVASIVE that used to take days in just seconds. 2 throughout the enterprise—in batch,

SAND TECHNOLOGY near real-time, and real-time modes, or PRICING: Vendor declined to provide on demand—to meet the analytical and SUN MICROSYSTEMS pricing. operational needs of the business. SYBASE Comprehensive auditing and reporting TERADATA capabilities enable business analysts to VERTICA support governance, risk and compli - ance initiatives. 2

ABOUT SEARCHDATA- PRICING: Vendor declined to provide MANAGEMENT .COM pricing.

METHODOLOGY

DATA WAREHOUSE PLATFORM PRODUCT DIRECTORY 11 KALIDO KOGNITIO

Kalido Dynamic Information WX 2 analytical database INTRO Warehouse (DIW)

SELECTING THE WX 2 is designed to be one of the fastest, RIGHT DATA Kalido’s Dynamic Information most mature analytical on WAREHOUSE PLATFORM Warehouse is designed to automate the market, producing answers to FOR YOUR ORGANIZATION the creation and lifecycle management queries in just seconds, and offering of enterprise data warehouses based a wide range of deployment options INDEX on your business model. s (software, appliance or DaaS). 1 s

1010DATA COMPANY WEBSITE: www.kalido.com/ COMPANY WEBSITE: www.kognitio.com ASTER DATA FOUNDED: 1997 FOUNDED: 2005 SYSTEMS SUMMARY: Kalido Dynamic Information SUMMARY: Kognitio WX 2 is designed to IBM Warehouse integrates data from multi - help midsized and large enterprises ILLUMINATE ple sources and powers operational and gain unprecedented analytical access SOLUTIONS strategic reporting. The software keeps to, and insight from, massive amounts GREENPLUM track of history so users can do as-of of data, thus allowing them to remain INFORMATICA and before-and-after time-based com - ahead of their competitors. Available

KALIDO parisons. DIW is built to handle busi - via traditional software license, as an ness change, adapting to deliver consis - integrated data warehouse appliance, or KOGNITIO tent information even during mergers through Kognitio’s DaaS, WX 2 enables MICROSOFT CORP. and acquisitions, reorganizations or users to reliably gain greater benefits in NETEZZA market consolidation. Its business less time by querying, in detail, vast ORACLE model-driven approach helps business amounts of granular data in just sec - CORPORATION and IT gather user requirements and onds. WX 2 can obtain results several PARACCEL prototype, iterate and deploy the data times faster than competing data ware - PERVASIVE warehouse quickly. DIW can save house appliances and up to 80 times

SAND TECHNOLOGY organizations significant deployment faster than typical software-only data - time and resources compared with cus - bases. SUN MICROSYSTEMS tom-built approaches, and it reduces SYBASE the manpower required to support PRICING: Kognitio WX 2 offers a flexible TERADATA operational implementations. 2 pricing model dependent on the deploy - VERTICA ment model, starting from as low as PRICING: Vendor declined to provide $30,000 per TB (software-only pricing pricing. DIW is sold only as part of the based on a 40 TB system). ABOUT SEARCHDATA- Kalido Information Engine. MANAGEMENT .COM

METHODOLOGY

DATA WAREHOUSE PLATFORM PRODUCT DIRECTORY 12 MICROSOFT CORP. NETEZZA

SQL Server 2008 AppNexus INTRO

SELECTING THE Microsoft SQL Server 2008 is a scalable AppNexus is built to allow users to RIGHT DATA and reliable data warehouse platform perform intensive processes, analytics WAREHOUSE PLATFORM for business intelligence. Through inte - and data warehousing faster and more FOR YOUR ORGANIZATION gration with the Microsoft BI platform, easily with data warehouse in the it is designed to enable actionable cloud. s INDEX business decisions at low cost. s COMPANY WEBSITE: www.netezza.com 1010DATA COMPANY WEBSITE: www.microsoft.com FOUNDED: 2000 ASTER DATA FOUNDED: 1975 SUMMARY: AppNexus has partnered with SYSTEMS SUMMARY: Microsoft SQL Server 2008 is Netezza to offer its customers a data IBM a scalable and reliable data warehouse warehouse cloud service. As a stand - ILLUMINATE platform for business intelligence. alone system, Netezza appliances are SOLUTIONS Through integration with the Microsoft designed to offer a fast and easy imple - GREENPLUM BI platform, it enables actionable busi - mentation process, and customers real - INFORMATICA ness decisions at the lowest cost. Take ize a fast time to value with Netezza

KALIDO advantage of a comprehensive Data implementations. The AppNexus Warehouse and BI platform in SQL “Netezza in the Cloud” service extends KOGNITIO Server. SQL Server 2008 ships with a the simplicity of Netezza implementa - MICROSOFT CORP. world-class relational database, analy - tions by offering rapid provisioning, NETEZZA sis services, reporting services and inte - scalability on demand, and pay-as-you- ORACLE gration services. Top new features go pricing based on utilization. This CORPORATION include: data compression, back-up gives customers a compelling new and PARACCEL compression, partitioned table paral - better option for their immediate, and PERVASIVE lelism, star join query optimizations, often elastic, terabyte-scale data analy - 2 SAND TECHNOLOGY resource management, change data sis needs. capture, MERGE SQL statements, and SUN MICROSYSTEMS scalable integration services. PRICING: Vendor declined to provide SYBASE pricing. TERADATA PRICING: Enterprise Edition: a comprehen - VERTICA sive data management and business intelligence platform providing enter - prise-class scalability, high availability, ABOUT SEARCHDATA- and security for running business-criti - MANAGEMENT .COM cal applications; Channels: Retail, VL, OEM, ISVR, SPLA; Processor Pricing: METHODOLOGY Retail $24,999, Example, $23,911 Server Plus CAL Pricing: Retail $13,969 with 25 CALs; Example, $8,487, $162 per additional CAL

DATA WAREHOUSE PLATFORM PRODUCT DIRECTORY 13 NETEZZA ORACLE CORPORATION

Data Warehouse Appliances HP Oracle Database Machine INTRO

SELECTING THE To eliminate the need for constant Oracle serves database and data RIGHT DATA tuning and the technology bottlenecks warehouse markets with products WAREHOUSE PLATFORM caused by slow disk transfer I/O rates, including Oracle Database 11g and FOR YOUR ORGANIZATION slow network transfer rates and ineffi - the HP Oracle Database Machine. s cient caching, Netezza developed a sys - INDEX tem designed specifically for analytical COMPANY WEBSITE: www.oracle.com processing on extremely large amounts FOUNDED: 1978 1010DATA of data. s SUMMARY: The HP Oracle Database ASTER DATA Machine is designed as a complete sys - SYSTEMS COMPANY WEBSITE: www.netezza.com/ tem, including software, servers, net - IBM FOUNDED: 2000 working and storage designed to run ILLUMINATE SUMMARY: Netezza’s massively parallel multi-terabyte data warehouses 10 SOLUTIONS architecture takes a different approach times faster than conventional data GREENPLUM to processing queries than traditional warehouse systems. At the heart of this INFORMATICA database architecture. Where standard system is the HP Oracle Exadata Stor -

KALIDO computing architectures target opera - age Server, which has smart storage tions on individual data elements, software that offloads data-intensive KOGNITIO Netezza architecture is all about query processing from database servers MICROSOFT CORP. “streaming” processing. Rather than closer to the data. The Database NETEZZA shuttling data between disk and memo - Machine runs Oracle Database 11g and ORACLE ry for processing once a query comes in, Real Application Clusters on Oracle CORPORATION data streams off the disk and through Enterprise , and also includes PARACCEL query logic loaded into a field program - complete Infiniband infrastructure. It’s PERVASIVE mable gate array (FPGA). The FPGA designed to be simple and fast to imple -

SAND TECHNOLOGY and processor, together with 400 GB of ment and cost-effective. The Database disk storage, reside on each of the mas - Machine can linearly scale storage SUN MICROSYSTEMS sively parallel nodes called snippet pro - capacity, processing power and network SYBASE cessing units (SPUs). Queries are opti - bandwidth as your data warehouse TERADATA mized across the SPUs for maximum grows. 2 VERTICA performance and power efficiency. 2 PRICING: Pricing can be found at PRICING: Vendor declined to provide http://www.oracle.com/corporate/ ABOUT SEARCHDATA- pricing. pricing/exadata-pricelist.pdf MANAGEMENT .COM Vendor declined to provide additional information. METHODOLOGY

DATA WAREHOUSE PLATFORM PRODUCT DIRECTORY 14 PARACCEL PERVASIVE

The ParAccel Analytic Pervasive Data Warehouse INTRO Database (PADB)

SELECTING THE Pervasive Data Warehouse is based RIGHT DATA PADB is a purpose-built data ware- on Pervasive Data Integrator and is WAREHOUSE PLATFORM housing and analytic database manage - designed to quickly scope, plan and FOR YOUR ORGANIZATION ment system that boasts load-and-go deliver an analytical data warehouse simplicity. PADB's innovative, MPP- to meet your business intelligence INDEX columnar design makes it the fastest objectives. s data warehousing platform on the 1010DATA planet, period. 1s COMPANY WEBSITE: www.pervasiveintegra - ASTER DATA tion.com SYSTEMS COMPANY WEBSITE: www.paraccel.com FOUNDED: 1982 IBM FOUNDED: 2005 SUMMARY: The Pervasive Data Ware - ILLUMINATE SUMMARY: The ParAccel Analytic Data - house delivers thorough documentation SOLUTIONS base is a robust, relational and fully- and a detailed design model for your GREENPLUM transactional DBMS that brings trans - data warehouse. An optional require - INFORMATICA formational speed to data warehousing, ments-gathering phase can be used to

KALIDO analytic and business intelligence (BI) define and design the integration applications. Its impressive speed is processes necessary for the ETL portion KOGNITIO built-in so that high-maintenance tun - of the project. The design can then be MICROSOFT CORP. ing structures like indexes, materialized implemented by your developers or by NETEZZA views and summary tables are not nec - a Pervasive consultant. Based on your ORACLE essary. This versatile performance is objectives, data warehouse consultants CORPORATION not dependent on a design will be able to design the analytical PARACCEL —our schema-neutral approach lets model for the data warehouse and rele - PERVASIVE you also unleash the power of ParAccel vant mart, identify the necessary source

SAND TECHNOLOGY for complex and ad hoc queries even on data elements required to meet busi - normalized and de-normalized schemas. ness objectives, and determine the nec - SUN MICROSYSTEMS Performance features include columnar essary SQL scripts to create the analyti - SYBASE data storage, massively parallel pro - cal model. Pervasive Data Warehouse TERADATA cessing, patent-pending SAN integra - requires Pervasive Data Integrator (see VERTICA tion, a patent-pending analytic query company website for details). 2 optimizer and scalable data loading. 2 PRICING: Vendor declined to provide ABOUT SEARCHDATA- PRICING: Vendor declined to provide pricing. MANAGEMENT .COM pricing.

METHODOLOGY

DATA WAREHOUSE PLATFORM PRODUCT DIRECTORY 15 SAND TECHNOLOGY SUN MICROSYSTEMS

SAND/DNA Access MySQL INTRO

SELECTING THE SAND/DNA Access provides data MySQL database is an open RIGHT DATA management software and best source database designed for fast WAREHOUSE PLATFORM practices for customers with large performance, high reliability, ease FOR YOUR ORGANIZATION amounts of data they want to access of use, and dramatic cost savings. 1 s and store. 1 s INDEX COMPANY WEBSITE: www.sun.com COMPANY WEBSITE: www.sand.com FOUNDED: 1982 1010DATA FOUNDED: 1983 SUMMARY: MySQL database is the world’s ASTER DATA SUMMARY: The SAND/DNA product suite most popular open source database SYSTEMS is designed to complement the data because of its fast performance, high IBM lifecycle management capabilities avail - reliability and dramatic cost savings. It ILLUMINATE able to organizations as their data has also become the database of choice SOLUTIONS warehouse tools grow in size. Users can for a new generation of applications GREENPLUM move infrequently used data into a built on the LAMP stack (Linux, INFORMATICA repository where it’s stored in a com - Apache, MySQL, PHP/Perl/Python).

KALIDO pressed format while remaining easily MySQL runs on more than 20 plat - accessible. Customers can quickly forms, including Linux, Windows, OS X, KOGNITIO respond to changing business needs, Solaris OS, and Netware, providing flexi - MICROSOFT CORP. regulatory compliance requests and bility. MySQL Enterprise powers a range NETEZZA changing market data without impact - of applications—from business-critical ORACLE ing their users and online warehouse. transactional systems to online eCom - CORPORATION As a result, infrastructure costs are merce, data warehousing, reporting and PARACCEL reduced and operational performance is analytics systems. Its “shared nothing” PERVASIVE improved, all without compromising architecture doesn’t require additional

SAND TECHNOLOGY users’ ability to access granular histori - infrastructure investment and provides cal data. SAND/DNA works with Ora - 99.999% data availability with no point SUN MICROSYSTEMS cle, Microsoft, DB2, SAS, SAP, Teradata of failure. SYBASE and Netezza. TERADATA PRICING: There is no license fee for VERTICA PRICING: DSAND/DNA Access pricing is MySQL or for MySQL Enterprise. How - based on the size of the flat file data ever, MySQL Enterprise offers various stored in SAND’s repository, and it is a levels of service and support, beginning ABOUT SEARCHDATA- perpetual license. at $599/year for Basic support. The var - MANAGEMENT .COM ious levels and their pricing are avail - able on the main MySQL website . METHODOLOGY

DATA WAREHOUSE PLATFORM PRODUCT DIRECTORY 16 SYBASE TERADATA

Sybase IQ 15 The Teradata Purpose-Built INTRO Platform Family

SELECTING THE Sybase IQ 15 is a highly optimized RIGHT DATA analytics server designed to deliver The Teradata purpose-built platform WAREHOUSE PLATFORM faster results for mission-critical busi - is designed for speed, scalability and FOR YOUR ORGANIZATION ness intelligence, data warehouse and affordability and is built to handle a reporting tools on any standard hard - high number of complex queries. s INDEX ware and operating systems. s COMPANY WEBSITE: www.teradata.com 1010DATA COMPANY WEBSITE: www.sybase.com FOUNDED: 1979 ASTER DATA FOUNDED: 1984 SUMMARY: Teradata’s database platforms SYSTEMS SUMMARY: Sybase IQ 15 is a column- are designed for scalability and afford - IBM based analytics server that is designed ability and are built for analyzing data ILLUMINATE to enable enterprises to turn raw data and processing the increasing volumes SOLUTIONS into actionable information through and growing complexity of queries at GREENPLUM analytics without the constant tuning the highest performance levels. The INFORMATICA required by some row-based databases platform family offers customers

KALIDO or the approach to business intelligence options that take full advantage of all typically taken by hardware data ware - the power of Teradata anywhere in the KOGNITIO house appliances. Sybase IQ 15 works enterprise—as an active data ware - MICROSOFT CORP. with diverse data—including unstruc - house, enterprise data warehouse, NETEZZA tured data—and diverse data sources entry-level data warehouse appliance, ORACLE and is designed to deliver unsurpassed special-purpose data mart, or sandbox CORPORATION query performance for the most com - environment. Teradata platforms are PARACCEL plex analytics tasks within an afford - recognized for speed, the amount of PERVASIVE able, eco-friendly and manageable envi - data that can be queried and the num - 2 SAND TECHNOLOGY ronment. ber and complexity of queries that can be run. SUN MICROSYSTEMS PRICING: Sybase IQ is available in three SYBASE editions: Sybase IQ Small Business Edi - PRICING: The Data Mart Appliance TERADATA tion starting at $120,000, which includes 551 starts at $56,000/terabyte; the VERTICA four CPUs; Sybase IQ Enterprise Edition, Extreme Data Appliance 1550 starts at at $64,800 per CPU; Sybase IQ Single $16,500/terabyte; the Data Warehouse Application Server Edition, starting at Appliance 2550 starts at $99,000/ter - ABOUT SEARCHDATA- $2,595 per CPU abyte; and the Active Enterprise Data MANAGEMENT .COM Warehouse 5555 starts at $150,000/ terabyte. http://www.teradata.com/t/ METHODOLOGY WorkArea/DownloadAsset.aspx?id=4 682

DATA WAREHOUSE PLATFORM PRODUCT DIRECTORY 17 VERTICA VERTICA

Vertica Analytical Database Vertica Analytic Database INTRO for the Cloud

SELECTING THE As businesses become more analytical RIGHT DATA to gain competitive advantage and Vertica Analytic Database for the WAREHOUSE PLATFORM comply with new regulations, data Cloud is an on-demand version of FOR YOUR ORGANIZATION warehouses are pushed to answer more Vertica’s grid-enabled columnar ad hoc questions from more people database hosted on ’s Elastic INDEX analyzing vastly larger volumes of data, Compute Cloud. The pay-as-you-go often in real time. The Vertica Analytic offering enables companies to create 1010DATA Database is built specifically to meet large, high-performance analytic data ASTER DATA this need. 1 s marts without up-front data center SYSTEMS costs and delays. 1 IBM COMPANY WEBSITE: www.vertica.com ILLUMINATE FOUNDED: N/A COMPANY WEBSITE: www.vertica.com SOLUTIONS SUMMARY: Vertica has these features and FOUNDED: N/A GREENPLUM functions in the latest edition: Column SUMMARY: Vertica Analytic Database for INFORMATICA orientation—query 50x-200x faster by the Cloud’s latest version features inno -

KALIDO eliminating costly disk IO; “scale-out” vations designed to allow Vertica to MPP architecture—scale limitlessly just manage terabytes of data faster and KOGNITIO by adding new servers to the grid; more reliably: “scale-out” grid architec - MICROSOFT CORP. aggressive data compression—reduce ture that handles changing workloads NETEZZA storage costs by up to 90%; automatic as elastically as the cloud; aggressive ORACLE high availability—run nonstop with data compression to keeps storage CORPORATION automatic replication, failover and costs low; and an automatic K-Safety PARACCEL recovery; and deployment flexibility— that provides replication, failover and PERVASIVE deploy on Linux or VMware or in the recovery in the cloud. Vertica for the

SAND TECHNOLOGY Amazon Cloud to handle a wide variety Cloud completely changes the econom - of projects . 2 ics of BI, making it possible to rapidly SUN MICROSYSTEMS initiate a much broader spectrum of SYBASE PRICING: Vendor declined to provide analytic projects and businesses, TERADATA pricing including ad hoc and short-lived busi - VERTICA ness analytic projects and new analytic SaaS businesses. 2

ABOUT SEARCHDATA- PRICING: Vendor declined to provide MANAGEMENT .COM pricing.

METHODOLOGY

DATA WAREHOUSE PLATFORM PRODUCT DIRECTORY 18 VERTICA

Vertica Virtualized INTRO Analytical Database

SELECTING THE RIGHT DATA The Vertica Virtualized Analytic WAREHOUSE PLATFORM Database is the first high-performance FOR YOUR ORGANIZATION analytic database to run as a virtual machine in private enterprise compute INDEX clouds. 1

1010DATA COMPANY WEBSITE: www.vertica.com ASTER DATA FOUNDED: N/A SYSTEMS SUMMARY: The Vertica Virtualized IBM edition combines a full-featured version ILLUMINATE of the Vertica Analytic Database with SOLUTIONS VMware’s Hypervisor virtualization GREENPLUM platform and an optimized self-con - INFORMATICA tained Linux . It is built

KALIDO as a simple-to-deploy, self-contained software package that runs on any KOGNITIO VMware-supported hardware. Using MICROSOFT CORP. VMware deployment tools, users can NETEZZA install any number of Vertica “virtual ORACLE appliances” and then provide any one CORPORATION of the instances with a list of the IP PARACCEL addresses of the others, and Vertica will PERVASIVE automatically unite them into a cluster.

SAND TECHNOLOGY A virtualized Vertica cluster can use direct-attached storage (DAS) or SUN MICROSYSTEMS network-attached storage (NAS). 2 SYBASE

TERADATA PRICING: Vendor declined to provide VERTICA pricing.

ABOUT SEARCHDATA- MANAGEMENT .COM

METHODOLOGY

DATA WAREHOUSE PLATFORM PRODUCT DIRECTORY 19 SearchDataManagement.com is a guide for information management and business intelligence professionals. With its combination of original articles, news, learning guides, expert advice, white papers, webcasts and customized research, Search - DataManagement.com offers a rich collection of insights on how to efficiently man - age the data supply chain. The site also offers tips on vendors and product selection, as well as expert advice.

Visit SearchDataManagement.com for:

q Independent content: Award-winning, vendor-independent news and analysis. E COM BE BER q Expert advice: Our Ask the Experts section features advice from some of the EM A M AY! leading authorities in the data management domain. TOD q Learning guides: Search our comprehensive library for useful how-to guides on any data management topic.

q Vendor-produced content: We hand-select white papers and webcasts to address the most relevant data management trends, issues and solutions.

Guide methodology: TechTarget has not evaluated the products listed and/or described in this Directory and does not assume any liabil - ity arising out of the purchase or use of any product described herein, neither does it convey any license or rights in or to any of the evalu - ated or listed products. TechTarget has prepared this Directory from sources deemed reliable (including vendors, research reports and cer - tain publicly available information). TechTarget has used good-faith efforts to indicate when content has been provided by a vendor and, in some cases, has removed what was deemed to be overt marketing language. TechTarget is not responsible for any errors or omissions in this Directory or for interpretations thereof, and expressly disclaims all war - ranties as to the accuracy, completeness or adequacy of all content contained herein. This disclaimer of warranty is in lieu of all warranties whether expressed, implied or statutory, including implied warranties of merchantability or fitness for a particular purpose. Information in this Directory is current as of Q3 2009. For more recent information, please check the vendors’ websites. The opinions expressed herein are subject to change without notice. ©2009 TechTarget, Inc. All Rights Reserved. Reproduction and distribution of this publication in any form without prior written per - mission is forbidden. TechTarget and the TechTarget logo are registered trademarks of TechTarget, Inc.; all other trademarks are the prop - erty of their respective companies. To compile this guide, our editorial team initially consulted research reports by major analyst firms covering the business intelligence market and contacted vendors about products reviewed by those firms. Editors also conducted additional Internet research and solicited feedback from our expert contacts. A notice about the project was posted on SearchDataManagement.com and listed regularly in our email newsletters. Vendors were invited to submit listings via a form on the website. For vendors that did not submit listings, our editorial team compiled listings by excerpting information from the vendor’s website. All entries, whether they were vendor-submitted or compiled by our team, were edited for length and clarity and to remove overt marketing language. In order to best assist our readers in assessing products, our editorial team attempted to obtain basic pricing information for all products in this directory—requesting information from vendors mul - tiple times via email. Vendors that did not respond, or refused to provide any pricing information, have this statement on their listings: “Vendor declined to provide pricing.” Collection of data for this directory took place during Q2 and Q3 2009. As with any directory of this kind, products and vendors may change substantially at any time. Though every effort was made to make this directory as complete and accurate as possible, there may be changes, errors, omissions or vendors in this market not included in this guide. Nothing in this guide should be construed as endorse - ments, professional suggestions or advice. This directory should be used simply as a resource. We strongly urge you to supplement this with your own research and to contact vendors for the most up-to-date information about their companies or products. It is our intent to update this directory annually, but that is subject to change.

DATA WAREHOUSE PLATFORM PRODUCT DIRECTORY 20