Computer resources and informatics

Computers play today such an important role in theoretical chemistry that it deserves at least few words. Actually, we owe a very important part of the emergence of theoretical chemistry to the high power and the low cost of nowadays computers.

Van Gunsteren and Mark [262] stated the quality of the simulated properties of a molecular system will depend on
i) the quality of the theory or model,
ii) the accuracy of the interatomic interaction function or force field,
iii) the degree of sampling, statistics and convergence reached in the simulation,
iv) the quality of the simulation software, and
v) how competently the simulation software is used.
The two last points depend on how capable is the theoretical chemist with the computer and the computer tools used.

The computational tools developed in this thesis need software of good quality that is nowadays available at no cost thanks to the GNU^A.1 initiative and many other initiatives that distribute their free software on the web. Most of this software can be installed under the GNU/Linux operative system which is also free and and it can be installed in a cheap and powerful personal computer. It means that we can have useful tools such as numerical and graphic libraries, different compilers, databases and many different kinds of software at very low cost. Obviously the production of results needs very powerful computational resources and sometimes the availability of efficient computer facilities requires a constant maintenance work which is also expensive.

In any case, it is out of discussion that a deep knowledge of the tools in informatics and computational science improves the efficiency and accuracy of the theoretical chemist work, not only to compute the numerical results but to manipulate, validate and analyze the data. I will give as example, two cases in which there are factors that can influence on the numerical accuracy of the computed results.