A Brief Introduction To In-memory Computing

In-memory computing is the style of computing which stores data primarily in the RAM or Random Access Memory, the part of memory used for processing the data in a computer system. This technique inhibits the storage of data on hard drives unlike the computer architecture, says a Gartner analyst Massimo Pezzini. He further adds that this would minimize the data access latency and would re-shape the primary utility of hard disks from storing data to capture overflowing data or for recovering lost data.

In-memory computing is becoming more popular as the need for processing tera- and peta-bytes of data for business utility is on the rise. This style of computing enables the processing and storing of large volume of data on the main memory i.e. RAM itself and need to store the data separately is vanished. As all the data is stored on the processing unit itself, so time to bring the data from a storage unit to the main processing memory and then processing it, is reduced. Usually, this whole process takes a very small amount of time calculated in micro-seconds. However, as the number of transactions grows to millions and even to billions, the response of the application grows to large value. In-memory computing brings the advantage of reduced calculation and processing time by storing all the data in the main memory.

One of the misunderstood notions of In-memory computing is its utility exclusively for analytical applications only. However, the fact is that in-memory computing has been there since 90's and most of the organizations including educational institutions, militaries, media, automobile, healthcare and many others are practicing in-memory computing in form or the other.

The dropping price of hardware units required to setup in-memory processing units (DRAM being the crucial one) is another contributing factor in the popularity of this concept. However, in comparison to the unit price of DRAM unit with conventional units it may seem expensive but after including the cost of associated factors like reliability, processing capability, performance and impact on business productivity, in-memory computing is an optimum solution in a long run.

Another great advantage of in-memory computing is the fact that business decision-makers are never left behind for making time-sensitive decisions. All the real-time data can be loaded directly onto the main memory and efficient patterns and values can be calculated in a short duration of time. This enables them to keep a track on instantaneous market conditions and make decisions in favor of business in no time.