Oracle Updates Exalytics for Analytics Scale
Oracle's new Exalytics engineered system provides more memory, but sticks with the same silicon.
Oracle first announced its Exalytics engineered server system for analytics in October of 2011. This week Oracle is delivering the first major update to the platform with the new Exalytics X3-4 system.
Paul Rodwick, VP of product management for Oracle Exalytics, explained to Enterprise Apps Today that the Exalytics system works with existing information management systems and both structured and unstructured data sources. Exalytics is an in-memory analytics system and will identify what data should be pulled into memory.
With the Exalytics X3-4, Oracle is doubling the main memory of the system to 2 TB of RAM. The new system also delivers 2.4 TB of flash storage and 5.4 TB of traditional spinning disk hard drive storage.
Though the Exalytics X3-4 provides more memory and storage for users, it doesn't include a refresh of the Intel Xeon chips. Rodwick noted that the new Exalytics X3-4 is powered by the same Intel Xeon processors that powered the first Exalytics X2-4 systems that started shipping in 2012.
As to why Oracle has not updated the processors on the Exalytics X3-4, Rodwick said there will be a continued evolution of Exalytics over time.
"Exalytics is designed as a middle tier engineered system to run the business analytics suite," he said. "It really uses main memory to solve the key challenges of analytics processing."
From a software perspective the Exalytics platform includes Oracle Business Intelligence Foundation Suite, Oracle Endeca Information Discovery, Oracle Essbase and Oracle TimesTen In-Memory Database. Oracle Essbase helps with scenario modeling as well as planning applications. Essbase came to Oracle from its acquisition of Hyperion in 2007.
As is the case with most of the systems in Oracle's engineered systems portfolio, the Exalytics run on Linux.
According to Rodwick, the key challenge for in-memory analytics is typically about determining what data should be put in memory.
"Typically enterprises have more corporate data than they can fit into memory," he said. "That's one of the great things about Exalytics, in that it works with your existing data sources and helps you to identify through some intelligent algorithms what is the best data to pull into memory."
Sean Michael Kerner is a senior editor at Enterprise Apps Today and InternetNews.com. Follow him on Twitter @TechJournalist.