Differences

This shows you the differences between two versions of the page.

--- start [2013/02/28 15:55] – daniel
+++ start [2019/03/06 18:49] (current) – daniel
@@ Line 1: / Line 1: @@
 ~~NOTOC~~
 ====== Welcome to the potfit wiki ======
+-----
-//potfit// is a free implementation of the force-matching algorithm to generate effective potentials from ab-initio data.
+//potfit// is a free implementation of the force-matching algorithm to generate effective
+potentials from ab-initio reference data.
-It was originally developed by Peter Brommer and is now maintained at the [[http://www.itap.physik.uni-stuttgart.de|ITAP]].
-For further information about the history and funding see [[about|About potfit - the ITAP Force-Matching Code]].
 ===== Features =====
-  * Fit empirical potentials with the force-matching algorithm
+  * Fit empirical potentials for molecular dynamics with the force-matching algorithm
-  * Tabulated or analytic potentials are available
+  * Support for tabulated, analytic and OpenKIM potentials
-  * Different interactions like pair potentials, EAM, ADP and electrostatics available
+  * These interactions are currently supported:
-  * Directly export the potentials to IMD, our high performance MD code
+    * pair potentials
+    * angular pair potentials (plain or with coulomb)
-===== Download =====
+    * EAM and two-band EAM
+    * ADP
+    * electrostatics (coulomb and dipole interactions)
+    * MEAM
+    * Stillinger-Weber
+    * Tersoff and modified Tersoff
+    * [[https://openkim.org/|OpenKIM]] models
+  * Directly export the potentials to [[http://imd.itap.physik.uni-stuttgart.de|IMD]] or [[http://lammps.sandia.gov|LAMMPS]].
-//potfit// is available as open source software released under the [[license|GPL]].
+===== Requirements =====
-The latest version can be downloaded [[download|here]].
+//potfit// is provided as C source code which requires the following:
+  * a Python runtime for the //potfit// build system
+  * a C99 compliant compiler (tested with GCC, Clang, and ICC)
+  * one of the following LAPACK implementations:
+    * Intel MKL
+    * Accelerate Framework on Mac OS
+  * an MPI implementation for parallel execution (tested with OpenMPI)