Skip to content
GitLab
Explore
Sign in
Primary navigation
Search or go to…
Project
P
PuReMD
Manage
Activity
Members
Labels
Plan
Issues
Issue boards
Milestones
Iterations
Wiki
Requirements
Code
Merge requests
Repository
Branches
Commits
Tags
Repository graph
Compare revisions
Locked files
Deploy
Releases
Model registry
Monitor
Incidents
Analyze
Value stream analytics
Contributor analytics
Repository analytics
Code review analytics
Issue analytics
Insights
Model experiments
Help
Help
Support
GitLab documentation
Compare GitLab plans
Community forum
Contribute to GitLab
Provide feedback
Keyboard shortcuts
?
Snippets
Groups
Projects
Show more breadcrumbs
SParTA
PuReMD
Graph
9c66d51eec8efc2845eebea9edd885dde3649cbc
Select Git revision
Branches
9
dummy-atom-support
energy_minim
main
default
protected
new_setup
pg-puremd-charge-solver-opt-sdsc-hackaton
qeq-nonzero-net-charge
restraint_support
tensorflow_update
tensorflow_update_v2
Tags
1
v1.0-rc1
10 results
You can move around the graph by using the arrow keys.
Begin with the selected commit
Created with Raphaël 2.2.0
11
Aug
10
7
6
5
1
29
Jul
22
6
1
29
Jun
25
24
23
17
12
5
2
1
12
May
10
7
4
30
Apr
20
15
13
6
10
Mar
5
3
2
26
Feb
19
10
17
Jan
16
8
7
6
18
Dec
11
10
6
4
3
25
Nov
8
4
29
Oct
25
24
20
17
11
10
8
6
3
24
Sep
23
12
10
9
5
29
Aug
17
6
4
1
30
Jul
27
12
21
Jun
12
10
3
30
May
29
28
27
24
23
5
4
30
Apr
18
15
13
18
Mar
17
14
12
6
4
1
28
Feb
27
25
21
20
19
18
14
13
7
6
29
Jan
28
27
26
25
23
21
18
15
10
9
8
6
5
4
3
2
31
Dec
28
27
23
22
21
19
18
17
16
15
13
12
10
7
4
30
Nov
27
26
25
23
22
12
9
8
7
6
5
3
29
Oct
25
24
22
21
15
6
4
1
28
Sep
22
20
19
17
15
13
12
11
8
7
6
5
30
Aug
28
26
22
19
15
31
Jul
15
14
Jun
30
May
29
28
24
23
22
21
18
16
15
14
30
Apr
19
17
15
11
3
2
27
Mar
26
22
21
13
12
1
26
Feb
23
20
19
18
15
14
13
12
8
7
4
2
1
29
Jan
26
25
24
17
8
30
Nov
27
21
3
26
Oct
20
19
16
23
Sep
19
10
6
27
Aug
25
24
23
22
21
26
Jul
25
24
16
15
11
22
Jun
20
17
13
7
6
26
May
1
19
Apr
18
27
Mar
22
18
27
Feb
26
25
23
21
16
11
23
Dec
8
6
5
3
25
Nov
24
22
21
16
13
Oct
11
9
Sep
8
7
6
4
3
1
31
Aug
26
23
18
7
6
5
3
2
29
Jul
28
25
20
1
30
Jun
28
27
21
20
18
17
16
14
12
9
8
7
3
31
May
30
19
18
16
13
26
Apr
23
18
12
25
Mar
13
11
24
Feb
17
16
2
1
29
Jan
28
26
14
PG-PuReMD: backport missing valence angle correction from sPuReMD.
PG-PuReMD: only copy host generated grid to device on reneighboring steps (after rebinning atoms and regenerating the grid). Fix small NVE integration bug (reallocation checks before copying new atom info to device).
PG-PuReMD: small clean-ups to QEq dual solver code.
PG-PuReMD: change CUDA kernel error checking to only make calls to cudaDeviceSynchronize when in debugging mode (non-debug builds thus have better performance with kernel asynchronous launch behavior in single default stream). Change file I/O to only flush when in debugging mode (non-debug builds allow the I/O operations to be grouped for better performance).
PG-PuReMD: add dual QEq BiCGStab solver.
PG-PuReMD: add support for symmetric, half stored format (SYM_HALF_MATRIX) of the sparse matrix for the charge model (add initialization routines and fix-up solver).
PG-PuReMD: rework custom reduction functions to use less shared memory and to have the correct thread and block counts. Fix bug in charge matrix initialization (under-allocated space previously). Split charge matrix and bonds/hydrogen bonds memory management routines to mirror the similar splitting of the initialization routines. Code clean-up related to half vs. full list and sparse matrix formats.
PG-PuReMD: change NaN floating point checks on energies to also include infinity values.
PG-PuReMD: improve performance by conditionally re-running initialization kernels within Cuda_Init_Forces based on out-of-memory conditions (for charge matrix, bonds list, hydrogen bonds list).
Tools: fix issue with merge for parse_results command (total line count error).
Tools: merge updated parse_results command with support for different run types.
Tools: merge updated parse_results command with support for different run types.
Tools: add PuReMD custom geometry replication to geo_tool.py. Remove older awk scripts. Fix issue with silica 6000 atom PDB file (CRYST1 lines not 70 characters).
Update .gitignore (restart files). Add missing updates to m4 files (CUDA).
Tools: update run_sim.py to generate more robust Slurm and Torque/PBS job scripts with the MPI+CUDA code (added extra option for specifying additional flags to be used with the command used to invoking the mpi+x code).
PG-PuReMD: fix performance logging code around GPU code. Rearrange header files to allow some preprocessor definitions to be defined via options in configure script. Rework Autoconf and Automake code to allow passing flags directly to nvcc compiler wrapper (for CUDA code). Enable C11 and C++11 standard targets during compilation.
PG-PuReMD: fix compilation error (remove variable). Fix issue with dual CG solver for QEq (local arithmetic in SpMV was incorrect). Revert CG solver convergence criterion to use norm of the preconditioned residual vector.
PG-PuReMD: small fix for initializtion routine performance logging (timers not reset after outputting to log file).
PG-PuReMD: corrections for utilizing a SAI preconditioner in MPI code (errors introduced during previous code merges).
PG-PuReMD: revert CG solver convergance criterion to used preconditioned residual norm. Re-enable dual charge solver for QEq.
PG-PuReMD: small corrections to performance logging code.
PG-PuReMD: small corrections to performance logging code.
PG-PuReMD: ensure that all processors log performance timings. Fix timing reduction.
PG-PuReMD: rework performance logging code in linear solver routines to avoid excessive MPI communications.
PG-PuReMD: adjust performance logging to compute mean timings across all processors. Other code clean-up.
PG-PuReMD: fix issue with certain data structures not being reallocated when the local number of atoms owned by a processor increases after exchanging messages with neighbor processors. Fix an issue where the MPI send message buffer may be overwritten by a received message (Coll). Small fix to not overwrite the program status return value with local function return values. Other code clean-up.
Tools: fix geometry file extension auto-detection for run_md_custom. Add missing restart-related control file keywords.
PG-PuReMD: fix out-of-bounds mmeory accesses and unitialized data usage in energy and force tabulation routines for van der Waals and Coulomb interactions.
PG-PuReMD: fix type mismatches for SAI preconditioner code.
PG-PuReMD: fix issue with some interaction lists not being initialized on first step of simulations from restarted runs. Fix typos in C++ code for utilizing C code (__cplusplus). Clean up C code for utilizing C++ code (unconditional extern's for unmangled member naming). Ensure that CUDA thread and block sizes are correctly set. Tweaks to memory transfers and allocation logic in integration code (GPU). Fix logic error with charge matrix allocation in GPU code. Other code clean-up.
Tools: fix geo_format for run_md_custom (geo file is mandatory, so the file type extension should always happen).
PG-PuReMD: add BiCGStab solver for GPU code. Corrections to CUDA block and thread sizes for some kernels. Fix some variables being used with uninitialized values.
Build: default to -O2 optimization.
PG-PuReMD: fix issue with charge solver preconditioner refactoring rate causes issues with reneighboring actions (preconditionering rate was previously coupled with reneighoring rate for SAI but this causes issues for Jacobi, etc.). Be more greedy with memory allocation sizes to decrease reallocation frequency (MPI buffers, etc.). More GPU code clean-up.
PG-PuReMD: fix MPI buffer allocations sizes. Ensure that nonblocking MPI messages have completed for each dimension before continuing. Rework reallocation checks in integration routines. Temporarily disable CUDA-aware MPI code paths (need to perform packing/unpacking first on device before handing off pointers). Other code clean-up.
PG-PuReMD: fix host-device transfers for charge solver code (SpMV data transfer sizes). Remove unused code. Other general code clean-up.
PuReMD: fix compile errors with newer compilers due to function prototype mismatches.
PG-PuReMD: re-write MPI code to dynamically allocate buffer sizes. Other general code refactoring.
PG-PuReMD: fix issue with divergent MPI_Reduce calls by multiple MPI processes (timing logging code). Add run-time MPI routine error checking. Fix issue with upper limit of hydrogen atoms allowed being hard-coded in GPU code (use dynamic memory allocation instead). Other code clean-up.
PG-PuReMD: clean-up MPI custom datatype initialization and error reporting. Avoid truncations in buffer size calculations. Other code clean-up.
Loading