Let me start with recent past or I would say past?

Traditional business intelligence tools and corporate performance management solutions typically use an IO-based processing technique that makes the compilation and analysis of data a slow process. It limits processing and speed from hardware resources that use extensive IO to read and write to and from the disk. This challenge can be frustrating for users and for IT staff because it reduces productivity and can overwhelm systems during peak periods.

Then came the In-memory processing era and BI solutions like Qlikview and Tableau. With InMemory Computing, these tools are able to load data into memory, and because the process of memory read/write is very efficient and fast as compared to IO-based processing technique, it provides superior performance. However, InMemory Computing requires a lot of memory to serve large data sets and a large user base, and that is an expensive hardware proposition.

And then, ElegantJ BI introduces Managed Memory Computing (MMC), that is the best, most balanced solution when compared to IO and InMemory computing, because it allows the Administrator to choose which data or cubes should reside in memory and which should reside on disk. The Administrator can allocate and prioritize data and hardware resources with a click of a mouse, on demand, without changing any complex configurations. This allows the business to give priority to critical data, or data that is more frequently accessed, or to data that is analyzed by important team members. Administrators can allocate less frequently used data, or less critical data, to IO operations. Based on the priority, the Administrator can decide resource allocation, and the business can manage and balance the priorities of data analytics with the available hardware resources to ensure the best user experience, without the expense of additional hardware.

There are several other advantages to using Managed Memory Computing (MMC) in Business Intelligence tools. The first is that the solution applies a level of aggregation, before putting this data into memory, so instead of working on row flat data, it works on pre-computed aggregated data, and thereby provides superior performance. The second is that MMC provides a meta data layer for business users, and users can relate this business meta data and taxonomy to day-to-day work, and use a simple drag and drop process to serve their Deep Dive Analytical needs, without technical assistance.

When considering a Business Intelligence and Corporate Performance Management Suite, many businesses consider only the features and functionality of the solution. To achieve a clear picture of cost and resources, the business must also consider the processing speed and the true accessibility of the data, as well as the value-add and price vs. performance ratio. In order to ensure that the business selects the best solution it must measure the value of the solution, and its speed and accessibility against the cost of hardware and IT resources. Managed Memory Computing offers a definite cost and performance advantage and is a critical business factor that should be added to the detailed requirements of every business intelligence project.

Original post: InMemory Vs Managed Memory Computing

Share Facebooktwittergoogle_pluslinkedin