1 GROMACS 2016.3 Release Notes
2 ----------------------------
4 This version was released on March 14, 2017. These release notes
5 document the changes that have taken place in GROMACS since version
6 2016.2 to fix known issues. It also incorporates all fixes made in
7 version 5.1.4 and several since.
9 Fixes where mdrun could behave incorrectly
10 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
12 Fixed mdrun with separate PME ranks hanging upon exit
13 """"""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""
14 A recent fix for another issue led to mdrun hanging while communicating
15 with PME ranks to coordinate end-of-run performance statistics.
19 Fixed handling of previous virials with md-vv integrator
20 """"""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""
21 These quantities get written to checkpoint files only for the Trotter
22 pressure-coupling integrators that need them, but they were being
23 copied in do_md for all Trotter integrators. This meant that an
24 appending restart of md-vv plus nose-hoover plus no pressure coupling
25 truncated off a correct edr frame and wrote one with zero virial and
26 wrong pressure. And in the same case, a no-append restart writes a
27 duplicate frame that does not agree with the one written before
32 Fixed an incorrect check that nstlog != 0 for expanded ensembles
33 """"""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""
34 The original version was accidentally reversed, causing it to
35 fail when nstlog was not equal to 0.
37 Fixes for ``gmx`` tools
38 ^^^^^^^^^^^^^^^^^^^^^^^
40 Fixed ``gmx tune_pme`` detection of GPU support
41 """"""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""
43 Fixed spacing in ``gmx tune_pme`` call to thread-MPI mdrun
44 """"""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""
46 Fixed minor issues in ``gmx traj -av -af``
47 """"""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""
48 Made the description of the xvg y-axis more useful. Also works for
53 Removed rogue printing to xvg file in gmx mindist
54 """"""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""
55 ``gmx mindist -xvg`` none is now adhered to, and printing is preceded by
60 Fixed bug in ``gmx solvate -shell`` if it yielded 0 SOL.
61 """"""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""
62 In the transition from genbox to solvate, some incorrect logic was
67 Corrected output of ``gmx do_dssp -sc``
68 """"""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""
69 This code has always written a probability, and not a percentage, so
70 fixed the label. It still fits within the expected 8-character field.
74 Improved documentation
75 ^^^^^^^^^^^^^^^^^^^^^^
77 Made several minor improvements to documentation and messages to users
78 """"""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""
79 Removed documentation of unimplemented ``gmx trjconv -clustercenter``.
81 Introduced system preparation section to user guide, to create
82 somewhere to document the use and limitations of vdwradii.dat.
83 Enchanced documentation of solvate and insert-molecules, similarly.
87 Documented that we now support AMD GCN on Mesa/LLVM
88 """"""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""
89 AMD GPUs using Mesa 17.0+ and LLVM 4.0+ run GROMACS using OpenCL.
91 Documented running Clang static analyzer manually
92 """"""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""
94 Portability enhancements
95 ^^^^^^^^^^^^^^^^^^^^^^^^
97 Enabled avx512 in the GROMACS FFTW build only if the compiler supports it
98 """"""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""
99 Enabling avx512 requires GCC 4.9 or newer or Clang 3.9 or newer. Since
100 we support compilers older than those, we can not afford to enable
101 avx512 in ``GMX_BUILD_OWN_FFTW=on`` unconditionally.
103 Worked around false positives in SIMD test from bug in xlc 13.1.5
104 """"""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""
105 atan2(0,0) should return 0.0, which the GROMACS simd implementation
106 does. However, since at least one compiler produces -nan for the
107 standard library version it's better to compare with the known
108 correct value rather than calling std:atan2(0,0).
112 Fixed compile with icc of ``GMX_SIMD=None``
113 """"""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""
114 ICC defines invsqrt in math.h