in-memory computing (IMC)
In- memory computing( IMC) is a computing concept that can significantly increase computing speed. The data volumes and computing requirements of many mobile applications, web-based projects, artificial intelligence, real-time analyses or high-speed transactions are increasing rapidly and can hardly be handled with hard disk-based computer architectures.
In-memory computing is a concept for increasing computing speed. In this concept, the main memory acts as a database. Data is read and written directly in the main memory. There is no need to transfer data to hard disks, solid-state drives or other storage media. The use of main memory, which has a storage capacity of several terabytes( TB) in the IMC concept, reduces access times by several orders of magnitude compared with hard disk-based architectures. This increases performance and computing speed and reduces latency. In this context, SAP cites several 1,000 times faster evaluation of its SAP HANA compared to databases for online transaction processing( OLTP).
With the IMC concept, a lot of data is kept in the main memory, which speeds up the data exchange between the main memory, program and processor. The concept dates back to the 1990s and was based on an extension of working memory that was used as a database. Processed data was not written to the hard disks, but remained in the cache that formed the memory extension.