Oracle releases its own in-memory applications in response to SAP's Hana strategy. How do they stack up?
Oracle announced 13 new in-memory applications on Tuesday that the company said will "change business dynamics to quickly discover growth opportunities, make smarter decisions, reduce corporate costs and accelerate time-consuming workloads."
These are the kinds of performance promises we're used to hearing from SAP, which for three years has been steadily rolling out its Hana in-memory database. SAP claims Hana delivers "stunning" performance improvements and "innovative, never-before-possible" applications. We've interviewed a handful of customers who back these claims up, and SAP says more than 400 customers have already or are in the process of rolling out the technology.
So has Oracle matched SAP's performance improvements with these new apps? Not likely, for technical reasons that I'll explain. But Oracle may succeed in muddying the competitive waters, blunting the competitive threat and delivering good-enough performance gains to keep Oracle customers satisfied.
The new in-memory applications that Oracle announced fall into two categories: net-new applications that weren't possible without in-memory acceleration, and new versions of existing apps that Oracle says warrant separate, in-memory versions for organizations interested in higher performance. All 13 will be released by the end of the year, according to Oracle applications executive Steve Miranda, and the first three are set for release in May.
The seven net-new applications "address problems that we really hadn't looked at before," said Miranda in an interview with InformationWeek. Oracle's in-memory capabilities, he said, have made it possible to take on these particularly data- and processing-intensive applications.
The net-new apps include JD Edwards EnterpriseOne In-Memory Sales Advisor, JD Edwards EnterpriseOne In-Memory Project Portfolio Management, Oracle Supply Chain Management (SCM) In-Memory Consumption-Driven Planning, Oracle SCM In-Memory Performance Driven Planning, Oracle SCM In-Memory Logistics Command Center, Siebel CRM In-Memory Policy Analytics and Siebel CRM In-Memory Next Best Action. The first three apps listed are the ones set to ship in May.
The six new in-memory alternatives to existing apps are Oracle E-Business Suite In-Memory Cost Management, PeopleSoft In-Memory Project Discovery, PeopleSoft In-Memory Labor Rules and Monitoring, PeopleSoft In-Memory Financial Allocations Analyzer, PeopleSoft In-Memory Financial Position Analyzer and Hyperion EPM In-Memory Virtual Close. Here, Oracle made logic and algorithm changes within the applications to take advantage of in-memory performance, Miranda said.
How hard will it be to switch from the legacy apps to these new in-memory alternatives? Think of it as deploying a new application rather than doing a simple upgrade, as "application configuration changes, data migration and data restructuring are involved," said Miranda.
Oracle's announcement stated that the new apps will exploit DRAM, flash memory and the InfiniBand networking used in Oracle's Engineered Systems "to run 10-20 times faster than commodity hardware" -- a comparatively modest performance increase claim and precious little insight into how the new apps will work. We're used to hearing much bolder performance-improvement claims from SAP and other in-memory proponents. But then, Oracle's approach to "in-memory" is often very different than the technologies others are talking about.
SAP Hana, for example, is an entirely in-memory database, with all data held in DRAM. In the case of Exadata, which is the Engineered System powering Oracle's new in-memory applications, Oracle can tap DRAM and flash-based cache, but it's not an in-memory database and the lion's share of data is still stored on spinning disks. The new apps, together with Exadata's intelligent caching capabilities, will have to make the most of the available DRAM and cache.
Comparing this to Hana performance, I happened to interview SAP customer Maple Leaf Foods this week. The Canadian food manufacturer was one of the earliest Hana customers, and it's accelerating many SAP BW-based analytic applications. Profit-and-loss analysis, for example, took 15 to 18 minutes before Hana, but it now takes 15 to 18 seconds, according to Michael Correa, Maple Leaf's VP, Information Solutions. That's 60X.
No, these aren't scientific, apples-to-apples comparisons of the same apps or even the same types of apps, but they fit a pattern of Oracle delivering solid, though not-stellar performance gains. At Oracle Open World in October, Mark Hurd said Exadata routinely delivers 100X performance gains, yet Deutsche Bank and Garmin, two customers who shared Exadata deployment details at the event, reported 50% (2X) improvements in performance.
Keep in mind that Oracle's new in-memory apps are aimed at Oracle apps customers, not SAP customers. It's an attempt to show these customers that they will be able to keep up with competitors who might be running SAP Hana.
Oracle's most direct response to Hana aimed at SAP customers is the combination of Exadata and Exalytics, both of which are certified to run SAP apps. Maple Leaf Foods, for one, continues to use Oracle database to run BW "for now," according to Correa, while using Hana as a "sidecar" to accelerate selected applications where performance really counts. While the economics of pure DRAM are becoming more attractive, they're not low enough, yet, for Maple Leaf to justify running everything in-memory, Correa said. That said, Hana, not Exadata, is Maple Leaf's choice for acceleration.
These are the economics that will ultimately determine the success of Hana and of Oracle's new in-memory applications. It's up to each company to determine just how much faster it needs to run and how much business benefit it can wring from performance improvement.
The Agile ArchiveWhen it comes to managing data, donít look at backup and archiving systems as burdens and cost centers. A well-designed archive can enhance data protection and restores, ease search and e-discovery efforts, and save money by intelligently moving data from expensive primary storage systems.
2014 Analytics, BI, and Information Management SurveyITís tried for years to simplify data analytics and business intelligence efforts. Have visual analysis tools and Hadoop and NoSQL databases helped? Respondents to our 2014 InformationWeek Analytics, Business Intelligence, and Information Management Survey have a mixed outlook.