OK, I've gotten some additional clarification...
NX Nastran comes in four versions:
1)
NX Nastran Desktop, which is included as part of the NX CAE modules and is therefore installed with NX and can use ONLY the NX FEM tools for it's pre- and post-processing needs. And since this can only run on a single desktop (hence its name) where NX is installed, while it's technically possible to access more than 8 cores, for all practical purposes, 8 cores is the limit.
2)
NX Nastran Basic, which is sold as a standalone product that can be used with either NX CAE, or most other CAD/FEM products, for pre- and post-processing and is designed to be installed on it's own dedicated server if desired. However, again while it technically can access more than 8 cores, anything past that provides very little added benefit, the same limitation as was the case with
NX Nastran Desktop.
3)
NX Nastran Advanced, which is an add-on to
NX Nastran Basic but since it includes support for the DMP (Distributed Memory Parallel Processing) you can run the solver on a cluster of networked CPU's where you can potentially have access to an unlimited number of cores or processors, but again, for all practical purposes, the limit is generally assumed to be 512.
4) And last but not least, is the
NX Nastran Advanced Nonlinear Solver Bundle which includes both the
NX Nastran Basic and
NX Nastran Advanced products along with a suite of highly specialized non-linear tools used for simulations involving contact, large deformation, large strain and nonlinear materials.
As previously stated, while all the versions of NX Nastran are SMP (Shared Memory Parallel Processing) which supports up to 8 cores on a single desktop/server, only the last two versions,
NX Nastran Advanced and
NX Nastran Advanced Nonlinear Solver Bundle supports DMP which will allow for the practical support or more than 8 cores simultaneously.
As for the
NX Nastran LP-64 versus
NX Nastarn ILP-64 issue, this appears to be simply an evolution of the software over time where it was rearchitected to better leverage modern memory managers which now allows for the utilization of vast amounts of physical memory on a single server. It's not so much tied to multiple processors as it is to getting the most out of the actual memory installed on each desktop/server. However, I have asked for further clarification as to when exactly did this transition from an
LP-64 to the
ILP-64 architecture take place.
Anyway, I hope this helps.
John R. Baker, P.E.
Product 'Evangelist'
Product Engineering Software
Siemens PLM Software Inc.
Industry Sector
Cypress, CA
Siemens PLM:
UG/NX Museum:
To an Engineer, the glass is twice as big as it needs to be.
