PRESS Development Platforms To Include AT&T Global Information Solutions and IBM Massively Parallel Solutions

Informix to Deliver Parallel Technology for Loosely Coupled Clusters and Massively Parallel Systems Development Platforms To Include AT&T Global Information Solutions and IBM Massively Parallel Solutions Commercial Parallel Processing Conference, Chicago, Ill. (Sept. 12, 1994) -- Informix Software, Inc., (NASDAQ:IFMX), the leading provider of parallel processing database technology, today announced that development plans are on track to deliver INFORMIX-OnLine Dynamic Server(TM) version 8.0, for loosely coupled clusters and massively parallel (MPP) systems. INFORMIX-OnLine Dynamic Server 8.0, Project XMP (eXtended Massively Parallel processing) is designed for hardware platforms that extend the range of Informix's support for parallel processing systems, beyond symmetric multriprocessor, to loosely coupled clusters and to massively parallel systems. General availability is scheduled to begin second half of 1995. Additionally, the company also announced plans to jointly develop its Project XMP Dynamic Server database technology on a variety of clustered and massively parallel systems, including AT&T Global Information Solution's System 3600, and IBM's Power Parallel SP2(TM) system. Development agreements are also in progress with additional hardware vendors. Informix's Project XMP implementation is aimed at several computing environments where customers need to process and manage very large amounts of data, such as online transaction processing (OLTP), decision support systems (DSS), data warehousing, and batch processing . "For the majority of the commercial market, uniprocessor and SMP systems will continue to meet customer needs, will be extremely cost-effective, and will scale well beyond the current level of 16-way processors," said Tim Shetler, vice president of product management at Informix. "Although today's market consists primarily of uniprocessor and symmetric multiprocessor systems, we see clustered and massively parallel systems growing in popularity as customers move off proprietary mainframes onto UNIX. "Informix is targeting a solution that is appropriate for each of these multiprocessor architectures, whether it be SMP, clusters or massively parallel. We are taking a long-term approach to ensure our customers can evolve their architectures and to give them the flexibility to choose the best solutions for their computing needs." Clustered and massively parallel systems will offer customers the potential for tremendous availability and scalability for online transaction processing and decision support applications. In these environments, more CPU resources can be added to the system, allowing customers to scale the system to add more users. With more added CPU resources, and the added functionality of Informix's parallel data query (PDQ) capabilities, customers will be able to gain tremendous query performance. These systems will also offer the key benefit of extremely high availability. If a system within a cluster fails, it will not affect the availability of the other systems in the environment. DSA Roadmap The Project XMP implementation of INFORMIX-OnLine Dynamic Server for loosely coupled and massively parallel systems will be Informix's third database server release based on its Dynamic Scalable Architecture(TM) (DSA) technology. The first release, version 6.0, which became available in December of 1993, provides customers with a multithreaded database server which exploits the capabilities of SMP and uniprocessor architectures and delivers breakthrough database scalability, manageability, and performance. Version 6.0 includes data replication for high availability, and parallel processing capabilities, such as index builds, sorts, and backup and recovery. The second release, version 7.0, also currently available, extends DSA to a full parallel implementation, including features such as complete Parallel Data Query, scan, join, sort, backup/restore, index builds and scans, table reorganizations, mass inserts, updates, deletes, and database partitioning. Version 8.0, or Project XMP, scheduled to be available beginning second-half 1995, will migrate version 7.0's complete set of functionality and parallelization capabilities to loosely coupled clusters and massively parallel systems. OnLine Dynamic Server And Parallel Processing OnLine Dynamic Server is based on Informix's next generation, multithreaded parallel processing database architecture, DSA. DSA includes Parallel Data Query (PDQ), a parallel processing technology that helps customers manage data volume loads, system administration needs, and performance restraints of enterprisewide OLTP, batch and decision support applications. OnLine Dynamic Server with PDQ provides ten to one-hundred times performance gains over conventional databases by breaking long-running database processing tasks into subtasks and then processes those tasks in parallel. This approach maximizes the power of each central processing unit (CPU) in a multiprocessor hardware system. DSA and PDQ are designed to exploit all the hardware capabilities available on the open systems market including uniprocessor, symmetric multiprocessor, loosely coupled clusters and massively parallel systems. About Informix Informix Software is the leading supplier of high performance, parallel processing database technology for open systems. The company's database servers are the number one choice of computer hardware manufacturers for publishing Transaction Processing Council (TPC) benchmarks for UNIX-based systems. Informix products also include application development tools for creating client/server production applications, decision support systems, and ad-hoc query interfaces, and connectivity software that allows information to be shared transparently from PCs to mainframes within the corporate computing environment. Informix corporate headquarters is based in Menlo Park, California. ### INFORMIX is a registered trademark, and Dynamic Scalable Architecture, and OnLine Dynamic Server are trademarks of Informix Software, Inc. All other companies mentioned or other names indicated by (R) or (TM) are registered trademarks their respective manufacturer. -- %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% William Daul Advanced Support INFORMIX SOFTWARE INC. 4100 Bohannon Dr. (415) 926-6488 - wk Menlo Park, CA. 94025 uunet!infmx!billd or billd@informix.com