Electrostatic forces and energies are one of the major components of the total energy of biological macromolecules. However, computing the electrostatic field distribution in systems made of biological objects immersed in water is not trivial task because of the large degree of freedom associated with the water phase, which in general limits the applications of explicit model to system with sizes smaller than several hundred Angstroms. This problem is avoided by applying continuum electrostatic approaches to deliver the potential distribution, considering that the water and macromolecules are two distinctive dielectric media. Here we report development and implementation in DelPhi of a Gaussian model for atomic densities and its usage to deliver a smooth dielectric function. The performance of the Gaussian DelPhi was benchmarked against solvation energies of small molecules obtained with explicit water simulations and very good agreement was found. The Gaussian DelPhi was also shown to perform much better than standard calculations in delivering the potential distribution is the reaction center protein. Furthermore we report a parallelized DelPhi, which speed is more than 100 times faster than the sequential DelPhi while delivering the same potential and energies up to the 15 position after the decimal point.