<div dir="ltr">thank you, Peter and Pavel, for the clarifications. </div><br><div class="gmail_quote"><div dir="ltr" class="gmail_attr">On Thu, 12 Dec 2019 at 14:06, Pavel Ondračka <<a href="mailto:pavel.ondracka@email.cz" target="_blank">pavel.ondracka@email.cz</a>> wrote:<br></div><blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left:1px solid rgb(204,204,204);padding-left:1ex">I concur.<br>
<br>
In general for the serial test case on modern CPU (avx2 instructions)<br>
your runtime should be around or below 30seconds for single thread.<br>
<br>
However as this is almost 10 years old mobile CPU with just avx<br>
instructions the total runtime of slightly above 1 minute is expected.<br>
<br>
Regarding the scaling even when not memory bound, I can get around 35%<br>
runtime compared to serial run with openBLAS (MKL scales slightly<br>
better). Small speedups could be probably gained with some work on HNS<br>
section (as this is the worst scaling part which we have more or less<br>
under control) but for the DIAG part we just depend on the BLAS/LAPACK<br>
to scale properly.<br>
<br>
If you have multiple k-points and your total memory permits it, its<br>
best to use k-point parallelization and use OpenMP just for lapw0 and<br>
mixer...<br>
<br>
Pavel<br>
<br>
On Thu, 2019-12-12 at 13:42 +0100, Peter Blaha wrote:<br>
> It is perfectly ok for your hardware.<br>
> <br>
> The cpu time is not so important for you, what counts is the WALL-<br>
> time <br>
> (this is the time it really takes until it finishes).<br>
> <br>
> You can see that Hamilt parallelizes fairly well (3.7 vs. 12.3<br>
> seconds / <br>
> speedup factor 3.3), HNS is not so good (3.8 vs. 8.8 s / factor 2.3)<br>
> and <br>
> DIAG is worse (23.2 vs. 48.2 / factor 2.1).<br>
> <br>
> Part of the reason that you can never see a factor of 4 is the slow <br>
> memory access, so when 4 cores do some calculations, they have to<br>
> wait <br>
> sometimes for data from the memory.<br>
> <br>
> On machines with more cores and a better memory bus, you will get<br>
> other <br>
> speed-ups, but basically no machine can use all cores with 100% <br>
> efficiency because of this limited memory access.<br>
> <br>
> <br>
> On 12/12/19 1:07 PM, Hemza wrote:<br>
> > Hi everybody:<br>
> > I just finished updating my wien2k installation to 19.1 with<br>
> > openMP <br>
> > support (linux (4.19.88), gfortran (9.2.0), openblas-lapack-openmp <br>
> > (0.3.7), fftw3 (3.3.8), libxc (4.3.4)), and patches from <br>
> > "<a href="https://github.com/gsabo/WIEN2k-Patches" rel="noreferrer" target="_blank">https://github.com/gsabo/WIEN2k-Patches</a>".<br>
> > I intend to use it for relatively small cases (less than 25<br>
> > atoms/unit <br>
> > cell). I run 'x lapw1' on the test_case.<br>
> > With OMP_NUM_THREAD=4 in bashrc:<br>
> > --------------------------<br>
> > $ x lapw1<br>
> > STOP LAPW1 END<br>
> > 113.876u 2.097s 0:31.36 369.7% 0+0k 424+37840io 2pf+0w<br>
> > $ grep HORB *output1*<br>
> > test_case.output1: TIME HAMILT (CPU) = 13.5, HNS = <br>
> > 12.6, <br>
> > HORB = 0.0, DIAG = 87.3, SYNC = 0.0<br>
> > test_case.output1: TIME HAMILT (WALL) = 3.7, HNS = <br>
> > 3.8, <br>
> > HORB = 0.0, DIAG = 23.2, SYNC = 0.0<br>
> > --------------------------<br>
> > <br>
> > and with OMP_NUM_THREAD=1 , I got:<br>
> > -------------------------------------<br>
> > $ x lapw<br>
> > STOP LAPW1 END<br>
> > 69.380u 0.339s 1:09.88 99.7% 0+0k 352+37848io 2pf+0w<br>
> > $ grep HORB *output1*<br>
> > test_case.output1: TIME HAMILT (CPU) = 12.0, HNS = <br>
> > 8.8, <br>
> > HORB = 0.0, DIAG = 48.1, SYNC = 0.0<br>
> > test_case.output1: TIME HAMILT (WALL) = 12.3, HNS = <br>
> > 8.8, <br>
> > HORB = 0.0, DIAG = 48.2, SYNC = 0.0<br>
> > ------------------------------------<br>
> > I do not feel i really understand the output and I do not know if<br>
> > this <br>
> > timing are good, so I eager to read your comments!<br>
> > <br>
> > My machine ('inix -dm' output)<br>
> > ------------------------<br>
> > System: Host: dojo Kernel: 4.19.88-1-lts x86_64 bits: 64<br>
> > Desktop: i3 <br>
> > 4.17.1 Distro: Artix rolling<br>
> > Machine: Type: Laptop System: ASUSTeK product: K53SD v: 1.0<br>
> > serial: <br>
> > <root required><br>
> > Mobo: ASUSTeK model: K53SD v: 1.0 serial: <root<br>
> > required> <br>
> > BIOS: American Megatrends v: K53SD.202<br>
> > date: 11/02/2011<br>
> > Battery: ID-1: BAT0 charge: 33.8 Wh condition: 33.8/59.4 Wh (57%)<br>
> > Memory: RAM: total: 7.57 GiB used: 4.84 GiB (63.9%)<br>
> > RAM Report: permissions: Unable to run dmidecode. Are<br>
> > you root?<br>
> > CPU: Quad Core: Intel Core i7-2670QM type: MT MCP speed: 849<br>
> > MHz <br>
> > min/max: 800/3100 MHz<br>
> > Graphics: Device-1: Intel 2nd Generation Core Processor Family <br>
> > Integrated Graphics driver: i915 v: kernel<br>
> > Device-2: NVIDIA GF119M [GeForce 610M] driver: nouveau<br>
> > v: <br>
> > kernel<br>
> > Display: x11 server: X.org 1.20.6 driver:<br>
> > intel,nouveau <br>
> > unloaded: fbdev,modesetting,vesa<br>
> > resolution: <xdpyinfo missing><br>
> > Message: Unable to show advanced data. Required tool<br>
> > glxinfo <br>
> > missing.<br>
> > Network: Device-1: Intel Centrino Wireless-N 100 driver: iwlwifi<br>
> > Device-2: Qualcomm Atheros AR8151 v2.0 Gigabit<br>
> > Ethernet <br>
> > driver: atl1c<br>
> > Drives: Local Storage: total: 2.05 TiB used: 1.45 TiB (70.8%)<br>
> > Info: Processes: 300 Uptime: 1d 1h 46m Shell: bash inxi:<br>
> > 3.0.26<br>
> > -------------------------<br>
> > <br>
> > regards<br>
> > <br>
> > _______________________________________________<br>
> > Wien mailing list<br>
> > <a href="mailto:Wien@zeus.theochem.tuwien.ac.at" target="_blank">Wien@zeus.theochem.tuwien.ac.at</a><br>
> > <a href="http://zeus.theochem.tuwien.ac.at/mailman/listinfo/wien" rel="noreferrer" target="_blank">http://zeus.theochem.tuwien.ac.at/mailman/listinfo/wien</a><br>
> > SEARCH the MAILING-LIST at: <br>
> > <a href="http://www.mail-archive.com/wien@zeus.theochem.tuwien.ac.at/index.html" rel="noreferrer" target="_blank">http://www.mail-archive.com/wien@zeus.theochem.tuwien.ac.at/index.html</a><br>
> > <br>
<br>
_______________________________________________<br>
Wien mailing list<br>
<a href="mailto:Wien@zeus.theochem.tuwien.ac.at" target="_blank">Wien@zeus.theochem.tuwien.ac.at</a><br>
<a href="http://zeus.theochem.tuwien.ac.at/mailman/listinfo/wien" rel="noreferrer" target="_blank">http://zeus.theochem.tuwien.ac.at/mailman/listinfo/wien</a><br>
SEARCH the MAILING-LIST at: <a href="http://www.mail-archive.com/wien@zeus.theochem.tuwien.ac.at/index.html" rel="noreferrer" target="_blank">http://www.mail-archive.com/wien@zeus.theochem.tuwien.ac.at/index.html</a><br>
</blockquote></div>