Performance testing of ABACUS #1948
Replies: 10 comments 10 replies
-
Version: v3.1.3 dataset1:ABACUS-PW (CG)
dataset2:ABACUS-LCAO (gen-elpa)
|
Beta Was this translation helpful? Give feedback.
-
Version: v3.1.4 dataset1:ABACUS-PW
dataset2:ABACUS-LCAO
|
Beta Was this translation helpful? Give feedback.
-
Version: v3.2.0 dataset1:ABACUS-PW
dataset2:ABACUS-LCAO
|
Beta Was this translation helpful? Give feedback.
-
Is this test using MPI or OpenMP? |
Beta Was this translation helpful? Give feedback.
-
Machine: |
Beta Was this translation helpful? Give feedback.
-
maybe a test run using OpenMP is also helpful. I am not sure how to choose |
Beta Was this translation helpful? Give feedback.
-
version: v3.2.3
version: v3.2.3
version: v3.2.3
version: v3.2.3
|
Beta Was this translation helpful? Give feedback.
-
@Satinelamp The above are some results of testing on v3.2.3, which is a reference. |
Beta Was this translation helpful? Give feedback.
-
Is there any tests of performance comparison between ABACUS and VASP? In my experience, even when I use LCAO basis, ABACUS is obviously slower than VASP. |
Beta Was this translation helpful? Give feedback.
-
@Satinelamp Hi! Can you share the output files of VASP. There are only input files in the zip file. |
Beta Was this translation helpful? Give feedback.
-
Since v3.1.3, along with the release of new version, we will report the performance testing.
The geometric mean of the inverse of total time is used as the metric, and the value of v3.0.0 is set to 1.0 as the baseline.
ks_solver
.ks_solver
.c32_m64_cpu ali
in Bohrium platform to do the calculationOMP_NUM_THREADS=1 mpirun -np 16 abacus
PW samples:
LCAO samples:
Beta Was this translation helpful? Give feedback.
All reactions