LANL hosted a virtual "1st annual" MCNP user symposium (https://custom.cvent.com/184F82DD73C94B5DA7BCE5DA8F257665/files/event/248345be235644959142f8f72c74ef71/57477a4c75d948428577eaecc61faacd.pdf) . There were 500+ attendees, and ~75 presentations (with 1/3 given by folks from LANL). Here's what Andreas Vogt and I found most interesting and applicable for the nuclear well-logging community:
- MCNP 6.3 should be coming out this fall
- New theory and user manuals
- Will require a Fortran 2018-compliant compiler. Current Intel and gfortran compilers work. Flang (the llvm fortran compiler) does not, nor does the Mac llvm-based compiler.
- The source code build system now uses cmake. I don't particularly love cmake, but it's better than hand-editing Makefiles, which was the old way to do things.
- HDF5 support (specifically HDF5 version 1.10) has been added, so there are now HDF5 output, mctal, meshtal, and EEOUT files, and I think MCNP can use an HDF5 input file, too. One significant advantage of HDF5 is