Differences

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

Link to this comparison view

Both sides previous revision Previous revision
Next revision Both sides next revision
running_molpro_on_parallel_computers [2021/05/20 08:23]
doll typo
running_molpro_on_parallel_computers [2021/06/02 08:43]
qianli [Running Molpro on parallel computers]
Line 7: Line 7:
 There are different GA implementation options (runtimes), and there are advantages and disadvantages for using one or the other implementation (see [[GA Installation]]). There are different GA implementation options (runtimes), and there are advantages and disadvantages for using one or the other implementation (see [[GA Installation]]).
  
-The Molpro binary is built with the ''--with-sockets'' GA runtimewhich requires pre-allocation of GA memory in some large calculations+Since Molpro 2021.2 the [[#disk option]] is used by default in single node calculation, in which case large data structures are simply kept in MPI files. 
-**Failing to preallocate sufficient amount of GA memory may lead to crashes or incorrect results, unless the [[#disk option]] is used.** +The behavior of previous versions can be recovered by the ''--ga-impl ga'' command line option. 
-It is therefore very important to read and understand [[#memory specifications]] before trying to run large scale parallel calculations with Molpro.+However''--ga-impl ga'' requires pre-allocation of GA memory in many calculations if the ''socket'' GA runtime is used, and failing to preallocate sufficient amount of GA memory may lead to crashes or incorrect results. 
 +Preallocating GA is not required with the ''mpi-pr'' runtime of GA or with the disk option. 
 ===== GA Installation notes  ===== ===== GA Installation notes  =====