How to make VASP calculations faster by selecting the right number of cores?

Описание: Kindly Click Here: https://bit.ly/2UtvbHE
How to make VASP calculations faster by selecting the right number of cores?
A frequent question that is always asked by VASP users is: “I have a cell with X-number atoms and Y-number electrons. How many compute nodes (or cores) should I choose for my calculation?” It is an important question because using too many cores is not always a way to speed up our calculations and in many cases applying too many cores will be inefficient and the result is fewer jobs completed in a given computer time allocation. That is why we need to learn how to optimize VASP performance by selecting the right number of cores in our different calculations. I have seen even many experienced VASP users apply only one scenario for all types of their calculations to make their calculations faster, which is not right and efficient and they do not consider the parameters such as the number of k-points and the number of bands.
VASP currently offers parallelization and data distribution over bands and/or over plane wave coefficients, in addition to parallelization over k-points. To obtain high efficiency on modern multi-core machines, it is strongly recommended to use all at the same time. The important note that we need to consider is that we need to know our system and what we are going to calculate. For example, the way of optimizing VASP performance for lattice relaxation is different from band structure calculations with large number of KPOINTs and the size of our system (the number of atoms, number of bands) is very important and should be considered. So, here, I will talk about the different scenarios of VASP performance optimization for different types of calculations.