diff --git a/Release/pnetcdf-1.14.0.tar.gz b/Release/pnetcdf-1.14.0.tar.gz new file mode 100644 index 0000000..10f6bf1 Binary files /dev/null and b/Release/pnetcdf-1.14.0.tar.gz differ diff --git a/Release_notes/1.14.0.md b/Release_notes/1.14.0.md new file mode 100644 index 0000000..9b7c7c9 --- /dev/null +++ b/Release_notes/1.14.0.md @@ -0,0 +1,84 @@ +[PnetCDF](https://parallel-netcdf.github.io) Version 1.14.0 Release Notes (November 11, 2024) +------------------------------------------------------------------------------ + +* New features + + Intra-node aggregation for write requests -- When the number of MPI + processes allocated to a compute node is large, this feature can + effectively reduce the communication congestion caused by an overwhelming + large number of asynchronous messages posted during the collective write + of MPI-IO. This new feature can be enabled by setting the PnetCDF I/O hint + 'nc_num_aggrs_per_node' to the desired number of aggregators per compute + node. The non-aggregators send their requests to the assigned aggregators, + and then the aggregators make aggregated requests to the file. + [PR #156](https://github.com/Parallel-NetCDF/PnetCDF/pull/156). + + Support MPI derived data types that are constructed from the large-count + derived datatype constructors introduced in MPI 4.0. + [PR #145](https://github.com/Parallel-NetCDF/PnetCDF/pull/145). + +* New optimization + + When running sequentially (i.e. number of processes is 1), PnetCDF calls + the MPI independent I/O functions and avoids calls to MPI_Barrier, + MPI_Bcast, and MPI_Allreduce. + [PR #149](https://github.com/Parallel-NetCDF/PnetCDF/pull/149). + +* Configure options changed + + The default has been changed to build both shared and static libraries. + [PR #143](https://github.com/Parallel-NetCDF/PnetCDF/pull/143). + +* Configure updates: + + Fix `pnetcdf-config` of reflecting the installation path when installation + is done by running command `make install DESTDIR=/alternate/directory` + which prepends '/alternate/directory' before all installation names. + [PR #154](https://github.com/Parallel-NetCDF/PnetCDF/pull/154). + +* New constants + + A new C macro `NC_FillValue` replaces `_FillValue` and thus `_FillValue` is + now deprecated This conforms with NetCDF4's change in its version 4.9.3 + release. [PR #153](https://github.com/Parallel-NetCDF/PnetCDF/pull/153). + +* New PnetCDF hints + + 'nc_num_aggrs_per_node' -- To enable the intra-node aggregation, this I/O + hint can set to a positive integral value, which indicates the desired + number of processes per compute node to be selected as the aggregators. + Setting it to 0 disables the aggregation, which is also the default mode. + [PR #156](https://github.com/Parallel-NetCDF/PnetCDF/pull/156). + +* Build recipes + + When using OpenMPI on Mac OSX, a link error may appear. The work around is + to add `LDFLAGS=-ld_classic` into the configure command line. Thanks to + Rui Chen for reporting and provide the solution. + [Issue #139](https://github.com/Parallel-NetCDF/PnetCDF/issues/139). + +* Updated utility programs + + none + +* Other updates: + + More document for comparing PnetCDF and NetCDF4 has been added to file + doc/netcdf4_vs_pnetcdf.md + [PR #152](https://github.com/Parallel-NetCDF/PnetCDF/pull/152) and + [PR #140](https://github.com/Parallel-NetCDF/PnetCDF/pull/140). + +* New example programs + + C/flexible_bottom.c and C/vard_bottom.c - These two examples construct MPI + derived data types using absolute memory addresses first and use MPI_BOTTOM + when calling the PnetCDF flexible APIs. + +* New programs for I/O benchmarks + + C/pnetcdf_put_vara.c -- + * This program writes a series of 3D variables with 2D block-block + partitioning pattern. Each variable is a record variable. + [PR #150](https://github.com/Parallel-NetCDF/PnetCDF/pull/150). + + C/netcdf_put_vara.c -- + * This sequential NetCDF-C program writes a series of 3D variables. Each + variable is a record variable. + * This program and `C/pnetcdf_put_vara.c` can be used to compare the + performance of NetCDF and PnetCDF when running sequentially, i.e. one + process. + [PR #150](https://github.com/Parallel-NetCDF/PnetCDF/pull/150). + +* New test program + + test/testcases/flexible_large_count.c - tests flexible APIs that use MPI + derived datatypes created by MPI large-count datatype constructors. + [PR #145](https://github.com/Parallel-NetCDF/PnetCDF/pull/145). + + diff --git a/index.html b/index.html index e8b7f58..215bb51 100644 --- a/index.html +++ b/index.html @@ -32,7 +32,7 @@

PnetCDF: A Parallel I/O Library for NetCDF File Access

News