Project

General

Profile

Roadmap

2021-refactoring

about 11 months late (03/31/2020)

Target to finish major refactoring for 2021 release

10%

33 issues   (2 closed — 31 open)

Related issues
Feature #2896: Python packaging
Feature #2967: GPU reallocateDeviceBuffer improvements
Feature #3295: Expand gmxapi.modify_input use cases.
Feature #3307: General interface for communication between simulation ranks
Feature #3311: GPU infrastructure development
Feature #3313: Introduce and use opaque types for the DeviceStream and DeviceContext
Feature #3318: Use wrappers for the GPU buffer copy/allocations
Feature #3328: Testing framework for task assignment
Feature #3374: SimulationInput abstraction

2021-infrastructure-stable

about 7 months late (08/15/2020)

changes to versions of things supported and tested should stabilize at least several weeks before the beta

10%

119 issues   (12 closed — 107 open)

Related issues
Feature #2816: GPU offload / optimization for update&constraits, buffer ops and multi-gpu communication
Feature #2817: GPU X/F buffer ops
Feature #2885: CUDA version of LINCS
Feature #2886: CUDA version of SETTLE
Feature #2888: CUDA Update and Constraints module
Feature #2994: Data flow topology in gmxapi 2020
Feature #2996: gmxapi execution model
Feature #3029: GPU force buffer ops + reduction
Feature #3032: Clean up dev-manual structure
Feature #3038: Improvements to MD plugin development environment
Feature #3117: Coulomb FEP PME on GPU
Feature #3140: Allow explicit input definition for gmxapi.operation function wrapper
Feature #3142: centralize and clarify GPU force buffer clearing
Feature #3147: gmxapi workflow checkpointing
Feature #3148: Roadmap for gmxapi filesystem interactions.
Feature #3152: Infrastructure and patterns for expressing public interfaces
Feature #3160: implement direct comm for different src/target memory spaces
Feature #3288: Use build system infrastructure instead of custom scripts to manage API levels.
Feature #3363: Improve FEP testing
Feature #3379: C++ API for simulation input and output
Feature #3393: Allow gmxapi.commandline_operation to use STDIN
Feature #3394: improve gmxapi error handling
Feature #3401: Policy and procedure for installed headers (public API)
Feature #3402: Clarify distinction between public API documentation and developer docs.
Feature #3417: Make modular simulator feature-complete
Feature #3423: Implement additional temperature and pressure control algorithms for modular simulator
Feature #3424: Implement stochastic dynamics / langevin integrator in modular simulator
Feature #3425: Implement rerun for the modular simulator
Feature #3426: Support acceleration in modular simulator
Feature #3427: Support freeze groups in modular simulator
Feature #3428: Implement SIMD version of modular simulator propagators
Feature #3429: Implement graph for modular simulator
Feature #3430: Implement NMR restraints for modular simulator
Feature #3431: Implement virtual sites for modular simulator
Feature #3432: Implement essential dynamics for modular simulator
Feature #3433: Decide how to handle multisim with modular simulator
Feature #3434: Support box deformation in modular simulator
Feature #3435: Implement pull for modular simulator
Feature #3436: Implement AWH for modular simulator

2021

about 2 months late (12/31/2020)

Annual major release

13%

68 issues   (8 closed — 60 open)

Related issues
Feature #868: Implement parallelization support to analysis framework
Feature #2224: Proposed feature: conditional stop
Feature #2710: Separate dvdl for each molecule (of couple-moltype) when running FEP
Feature #2774: Refactor shell code into its own integrator
Feature #2806: SIMD algorithms for ARM SVE // nobnonded cluster and otheres
Feature #2890: GPU Halo Exchange
Feature #2934: GPU X Buffer ops
Feature #2944: Roadmap for thermostats / barostats in new propagation/integration scheme
Feature #3114: Improve GPU update-constraints module
Feature #3277: Allow testing feature that is partly implemented
Feature #3283: Support for the string method with swarms of trajectories in GROMACS
Feature #3286: Optionally skip initial coordinates from being written to output coordinates
Feature #3350: GPU update-constraints performance
Feature #3351: GPU update-constraints code organization and integration
Feature #3352: GPU update-constraints feature support
Feature #3371: Control FEP using AWH
Feature #3410: Random Acceleration Molecular Dynamics (RAMD)

future

Placeholder for issues to address in a not-yet-decided/not-yet-planned future release

8%

177 issues   (13 closed — 164 open)

Related issues
Feature #438: New tool: g_correl
Feature #720: permit pdb2gmx to choose a kind of HIS based on proton position
Feature #742: Enhancing the performance of the free energy code
Feature #760: Implement rigid body groups
Feature #889: pressure annealing
Feature #895: g_rmsf previous frame as reference
Feature #907: Add tests for verifying installed headers
Feature #951: Multiple versions of Gromacs (e.g., single and double) in the same library/binary
Feature #1053: L-BFGS doesn't use constraints or call do_em_step
Feature #1104: Implement dihedral restraints
Feature #1120: Make build work with multi-cofiguration generators
Feature #1137: Proposal for integrator framework (do_md) in future GROMACS
Feature #1142: Synchronizing the differences between point estimation and histogram estimation for free energies
Feature #1162: Implement gb_saltconc
Feature #1165: Multi-SIMD binaries
Feature #1167: trjconv -center would be more intuitive if it centered the center of mass, not the midpoint between minimum and maximum values
Feature #1168: g_density does not do what users think it does, given PBC over z and constant pressure simulation
Feature #1182: improve trajectory writing to support parallel I/O
Feature #1221: More generic position mapping for selections
Feature #1247: fix hardcoded references to atom names in analysis tools
Feature #1335: Improved interaction of free energies with pull code to better support Hamiltonian replica exchange with umbrella sampling
Feature #1347: future of tables
Feature #1397: gmx eneconv -offset 1.998 -dt 2 is broken for large times
Feature #1437: Online help formatting improvements
Feature #1464: implement PP-PME re-balancing
Feature #1518: Enable automatic build checking for newer versions
Feature #1625: Gromacs Python API
Feature #1665: improve free energy non-bonded kernel performance
Feature #1666: new approach for Verlet-scheme kernel generation
Feature #1715: improve cycle counting GPU sharing and multi-sim
Feature #1720: Manage GROMACS JIT caching
Feature #1782: OpenCL runtime/compiler version should be reported in the log...
Feature #1877: Support reading 3d-grid data from cryo-electron microscopy tomography and X-ray crystallography
Feature #1953: use more regular polymorphism for GPU code
Feature #1966: Hydrogen mass repartiniong
Feature #1972: external potential modules for refinement against experimental data
Feature #1994: explore using JIT for CUDA
Feature #2005: Full Correlation Analysis (FCA) as Trajectory Analysis Module
Feature #2034: Unit tests for bonded forces
Feature #2054: PME on GPU
Feature #2068: Access to low level classes
Feature #2070: Physical validation testing
Feature #2097: Move output of dhdl file entirely to edr, rather than having a separate text dhdl file
Feature #2132: Intermediate code for xvg handling
Feature #2137: Preliminary refactoring of constraints and update machinery.
Feature #2186: Potential change for logical improvements: move control of constraints purely to the .mdp
Feature #2229: Full Object Oriented Modularization of GROMACS MDRUN Codebase
Feature #2282: Density map toolset
Feature #2310: Let mdrun dump coordinates with non-finite energy
Feature #2354: develop configuration file support for control of task layout
Feature #2426: Trajectory analysis tool for functional mode analysis by partial least squares fitting (PLS-FMA)
Feature #2491: Allow constant bias AWH simulations
Feature #2545: Should grompp fix periodicity of input files?
Feature #2547: Add option of averaging and obtaining error bars for AWH PMFs
Feature #2574: iForceSchedule Abstraction
Feature #2579: Add dimensionality option to gmx trjconv -center
Feature #2581: User interface for hybrid Monte Carlo
Feature #2594: Multi-level GMX API
Feature #2601: Free energy calculations, soft-core potential
Feature #2622: Reduce severity of Berendsen warning
Feature #2715: Avoid requesting the user to recompile gromacs for Intel OpenCL support
Feature #2791: Implement a module for lambda-dynamics simulations (lambda_site module)
Feature #2840: Test that functionality does not compile that is not supposed to.
Feature #2866: Alternative non-bonded potentials
Feature #3069: Iterator over span and indices into this span (zip-style iterator)
Feature #3172: QM/MM Interface with CP2K
Feature #3285: Run simulations from the same tpr file with different random seeds