<![CDATA[Multiwfn forum / Calculation speed of ESP has been significantly improved!]]> - //www.umsyar.com/wfnbbs/viewtopic.php?id=360 Sat, 04 Jul 2020 01:12:50 +0000 FluxBB <![CDATA[Calculation speed of ESP has been significantly improved!]]> //www.umsyar.com/wfnbbs/viewtopic.php?pid=1229#p1229 Dear users,

An efficient code for evaluation electrostatic potential (ESP) was provided by Jun Zhang (author of LIBRETA electron integral library) recently and has been incorporated into Multiwfn since the version released on 2020-Jul-4.

Below is comparison of computational cost of calculating RESP charge for a medium-sized organic system (36 atoms, def2-TZVP basis set)

1.png

As can be seen, the new ESP evaluation code is significantly faster than the old one. If the number of available CPU cores is 30~40, the speed is even comparable to the case of invoking cubegen utility in Gaussian.

Now the rule of calculating ESP in Multiwfn is:
1 If fch/fchk/chk is used as input file and "cubegenpath" has been correctly defined in settings.ini, then cubegen will be invoked to evaluate ESP in most kinds of analyses (see Section 5.7 of Multiwfn manual for detail).

2 For other cases, the new ESP evaluation code is utilized to calculate ESP (However, if you want to switch to the old one, set "iESPcode" in settings.ini to 1).

]]>
Sat, 04 Jul 2020 01:12:50 +0000 //www.umsyar.com/wfnbbs/viewtopic.php?pid=1229#p1229