[Wien] compilation problem

[SANDEEP ARORA]

Dear wien2k user
        i have installed the wien2k 21.1 in a laptop with following
configuration
CPU: 8-Core 11th Gen Intel Core i7-11800H (-MT MCP-) speed/min/max:
4265/800/4600 MHz Kernel: 5.11.0-40-generic x86_64
Up: 1h 28m Mem: 2414.9/15743.8 MiB (15.3%) Storage: 953.87 GiB (5.2% used)
Procs: 335 Shell: bash 5.0.17 inxi: 3.0.38
with intel one api compiler without any compile time error.
on testing calculations with test_case, the following results are obtained

OMP_NUM_THREADS=1
18.195u 0.306s 0:18.52 99.8% 0+0k 0+37840io 1pf+0w
test_case.output1:       TIME HAMILT (CPU)  =     2.3, HNS =     1.8, HORB
=     0.0, DIAG =    14.1, SYNC =     0.0
test_case.output1:       TIME HAMILT (WALL) =     2.3, HNS =     1.8, HORB
=     0.0, DIAG =    14.1, SYNC =     0.0

OMP_NUM_THREADS=2
LAPW1 END
20.108u 0.367s 0:10.35 197.6% 0+0k 1064+37824io 9pf+0w

grep HORB *output1*
test_case.output1:       TIME HAMILT (CPU)  =     2.5, HNS =     2.3, HORB
=     0.0, DIAG =    15.4, SYNC =     0.0
test_case.output1:       TIME HAMILT (WALL) =     1.3, HNS =     1.1, HORB
=     0.0, DIAG =     7.7, SYNC =     0.0

OMP_NUM_THREADS=4
27.836u 0.527s 0:07.26 390.4% 0+0k 0+37824io 1pf+0w

TIME HAMILT (CPU)  =     3.3, HNS =     4.0, HORB =     0.0, DIAG =
 20.7, SYNC =     0.0
       TIME HAMILT (WALL) =     0.8, HNS =     1.0, HORB =     0.0, DIAG =
    5.2, SYNC =     0.0

  OMP_NUM_THREADS=8
       46.376u 0.964s 0:06.13 772.1% 0+0k 64752+37824io 457pf+0w

       test_case.output1:       TIME HAMILT (CPU)  =     5.7, HNS =
8.0, HORB =     0.0, DIAG =    33.1, SYNC =     0.0
test_case.output1:       TIME HAMILT (WALL) =     0.7, HNS =     1.0, HORB
=     0.0, DIAG =     4.1, SYNC =     0.0



with k-point parallelization
localhost k=2 user=39.923 wallclock=40.6
40.132u 0.797s 0:21.70 188.5% 0+0k 1344+65888io 15pf+0w


ocalhost k=4 user=99.881 wallclock=101.23
100.227u 1.461s 0:26.86 378.5% 0+0k 0+121024io 5pf+0w

localhost k=8 user=357.058 wallclock=360.6
357.865u 3.802s 0:47.47 761.8% 0+0k 0+230608io 10pf+0w




I want to know why there is almost the same speed with k-point
parallelization and omp (5.93 s/k-point vs 6.13s/k-pont). is it due to my
hardware limitations or i have done something wrong during compilation.
wien2k_options file is attached herewith.

Thanks
Sandeep Arora
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://zeus.theochem.tuwien.ac.at/pipermail/wien/attachments/20211127/6d2eed50/attachment.htm>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: WIEN2k_OPTIONS
Type: application/octet-stream
Size: 1091 bytes
Desc: not available
URL: <http://zeus.theochem.tuwien.ac.at/pipermail/wien/attachments/20211127/6d2eed50/attachment.obj>