New Programming Language Is Expected To Quadruple The Speed In Big Data

1it enterprise

1it enterpriseMIT announce a new programming language which is targeted to remedy the problem of memory management which can be a challenge for the conventional datasets but the entry of Big Data entering the picture can change the scenario. As per the current status of the programming language, it is providing a 4X speed boost on the common algorithms. The memory management is governed by the principle of locality in most of the computers today. What it means is the neighboring chunks of programming data is also needed in order to access the particular program data from a location of the memory which is not always the cane in Big Data. Instead, only a few bits of data from the data scattered across the huge data set is what the act on.

 

Fetching the data from main memory more frequently can slow down the execution considerably as it is the main performance bottleneck in case of today’s chips. This process can be better explained using an example like every time you need a sip of coffee you have to open the cabinet take bottle coffee beans and grind them, make coffee and have it.

 

The new language is named MILK which allows the application developers to manage memory in a much better way. OpenMP is an API like C and Fortran which helps in writing code for multicore processors is benefited by the MILK language.