In-Memory Simulation Virtual Cubes for Rapid What-If Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Business systems face challenges in performing advanced analytics and simulations, such as 'what-if' analysis, due to the computational demands of handling large datasets, which can slow down user interactions and potentially contaminate actual data during simulation processes.
Innovation Solution
The method involves forming virtual and online data cubes in in-memory storage, using a multi-dimensional schema like star or snowflake, to rapidly process simulation data without affecting actual online data, allowing for rapid execution and separate storage of simulation results, and enabling users to interactively change variables and view impacts without intermediate persistent storage.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Speed
If simulations are performed using traditional database storage and processing methods, then computational accuracy is maintained, but processing speed decreases and user interaction becomes slow
Solution Approach 1:
The patent creates virtual cubes that are copies of the actual online data cubes, allowing simulations to be performed on replicated data structures in memory without affecting the source data. This copying approach enables rapid simulation execution while maintaining data integrity and accuracy.
Solution Approach 2:
The patent changes the storage parameter from traditional disk-based databases to in-memory data structures (virtual and online data cubes). This parameter change from persistent storage to volatile memory dramatically improves processing speed and simulation execution efficiency while maintaining computational accuracy.
2Ease of operation
If simulation data is stored together with actual online data, then data integration is simplified, but data contamination and accuracy loss occur
Solution Approach 1:
The patent segments data into distinct virtual cubes (for simulation) and online data cubes (for actual business data), both stored in memory. This segmentation allows simulations to run on isolated copies while maintaining easy integration through the shared memory structure, preventing data contamination while preserving operational simplicity.
Solution Approach 2:
The virtual cube acts as an intermediary between simulation operations and actual online data. It provides a protective layer that allows simulations to access and manipulate data representations without directly affecting the source data, thus maintaining both ease of operation and data reliability.
3Adaptability or versatility
If repeated simulations are performed with variable changes, then analytical depth increases, but processing time accumulates and user experience deteriorates
Solution Approach 1:
The patent loads actual online data into virtual cubes in advance, before simulations are executed. This preliminary action of pre-loading data into memory eliminates the need for repeated disk I/O operations during simulation re-executions, allowing rapid variable changes and repeated simulations without accumulating processing time delays.
Data Source
AI summary
Methods and apparatus, including computer program products, are provided for in-memory simulations. The method may include forming a virtual cube including at least one change value representative of a change to a simulation; forming another cube including online data; and combining the virtual cube and the other cube to form a union cube representative of a result of the simulation, the virtual cube, other cube, and the union cube comprised in an in-memory storage and shared for at least a session associated with the simulation. Related systems, methods, and articles of manufacture are also disclosed.