deletions | additions       diff --git a/Parallelization.tex b/Parallelization.tex  index 9325ee8..8e77580 100644  --- a/Parallelization.tex  +++ b/Parallelization.tex 
...
structure codes must execute efficiently in modern computational  platforms. This implies support for massively parallel platforms and  modern parallel processors, including graphics processing units  (GPUs). Octopus has been shown to perform efficientlyboth  on parallel supercomputers~\cite{Andrade_2012,Alberdi_2014}. Octopus supercomputers with scaling to hundreds of thousands of cores~\cite{Andrade_2012,Alberdi_2014}. The code  also has a very efficient an  implementation of GPU acceleration~\cite{Andrade_2012,Andrade_2013}. acceleration~\cite{Andrade_2012_gpus,Andrade_2012} that has shown to be competitive in performance with Gaussian DFT running on GPUs~\cite{Andrade_2013}.  Performance is not only important for established methods, but also  for the implementation of new features. Ideally, developers should be  isolated as much as possible from the optimization and parallelization  requirements. The simplicity of real-space grids allows us to provide Octopus developers with building blocks that they can use to produce highly efficient code without caring about the details of the  implementation. In most cases, this building blocks allow developers to write  code that is automatically parallel, efficient, and that can transparently