Hello,
I would like to assess the performances of NEST on different hardwares. And I have some questions if I may.
I am thinking about reusing one of the benchmarks here : https://github.com/compneuronmbu/nest-benchmarks.
Let's assume we have two nodes : A and B. A has N cores and B has M cores. I want to know if NEST runs faster for a particular test case on A or B. Hence I am fully filling the nodes with MPI ranks and OpenMP threads. If my understanding is correct, I need to set NVP = MPI ranks x OpenMP threads hence NVP = N on A and NVP = M on B. At this stage, for a lot of applications I would just say "The application runs faster on A or on B." . What worries me for NEST is that N and M might be different, hence different NVP for the simulation on A and B. From the documentation it says that for different NVP we have different results. So it gets me a bit worried to say that NEST is faster on A or B for a particular test case if I am not computing the same thing. I was wondering if you could confirm whether this is (or not) an issue and if you could give me some guidelines as regards benchmarking NEST please. That would be fantastic. This is also problematic if I want to do a strong scaling study on a single node for instance.
I could run NEST on the smallest common multiple of N and M (let's call it P). This would mean P/N nodes of A and P/M nodes of B. I guess this would make sense for NEST. But this is not a single node comparison, this is more a core vs core comparison.
Hope this makes sense. And please feel free to correct me if I said something wrong. I do not have a lot of experience with NEST.
Best regards,
Conrad Hillairet.
Hi,
I have proposed the use of StackOverflow as a channel for questions. What
does everyone think about it if I transcribe the mailing list questions and
possibly their answers onto StackOverflow?
(For fellow SO'ers, I will be making them Community Wiki questions as not
to unfairly absorb all that future reputation 💸💰💵)
Dear everyone,
in today's meeting [1] we discussed the relations between the currently
open pull-requests [2], and found that we would need a separate meeting
to discuss some in more detail. To find an appropriate date I've put a
few random slots into a poll [3]. Please fill your availability if you
are interested to join, till tomorrow evening:
--> https://nuudel.digitalcourage.de/zreAHis9t4BPC5Rl
Anyone involved with any of the topics or interested to help
testing/proof-reading/…, is very welcome to join. Feel free to have a
look! Some discussions will – by their nature – however be quite
developer oriented.
I'll send connection details to participants when a suitable slot is found.
Best,
Dennis
[1]:
https://github.com/nest/nest-simulator/wiki/2021-03-01-Open-NEST-Developer-…
[2]: https://github.com/nest/nest-simulator/pulls
[3]: https://nuudel.digitalcourage.de/zreAHis9t4BPC5Rl
--
Dipl.-Phys. Dennis Terhorst
Coordinator Software Development
Institute of Neuroscience and Medicine (INM-6)
Computational and Systems Neuroscience &
Theoretical Neuroscience,
Institute for Advanced Simulation (IAS-6)
Jülich Research Centre, Member of the Helmholz Association and JARA
52425 Jülich, Germany
Building 15.22 Room 4004
Phone +49 2461 61-85062
Fax +49 2461 61- 9460
d.terhorst(a)fz-juelich.de
-------------------------------------------------------------------------------------
Forschungszentrum Juelich GmbH
52425 Juelich
Sitz der Gesellschaft: Juelich
Eingetragen im Handelsregister des Amtsgerichts Dueren Nr. HR B 3498
Vorsitzender des Aufsichtsrats: MinDir Volker Rieke
Geschaeftsfuehrung: Prof. Dr.-Ing. Wolfgang Marquardt (Vorsitzender),
Karsten Beneke (stellv. Vorsitzender), Prof. Dr.-Ing. Harald Bolt
-------------------------------------------------------------------------------------
Dear NEST Developers!
With the merge of #1843 a few moments ago, we have implemented a significant change in the way NEST documentation is built (thank you, Jochen!). Please merge the current master in all open PRs to avoid complications.
With the new setup, documentation can be built with `make doc` and will be built in the build directory and can also be installed. To build the documentation, you will need a range of Python packages. Please see doc/requirements.txt for what is needed.
We will continue to make improvements to the documentation build system, but these will be minor.
If you have any open PRs, please merge the current master as soon as possible into your branches.
Best regards,
Hans Ekkehard
--
Prof. Dr. Hans Ekkehard Plesser
Head, Department of Data Science
Faculty of Science and Technology
Norwegian University of Life Sciences
PO Box 5003, 1432 Aas, Norway
Phone +47 6723 1560
Email hans.ekkehard.plesser(a)nmbu.no<mailto:hans.ekkehard.plesser@nmbu.no>
Home http://arken.nmbu.no/~plesser
Dear NEST Users & Developers!
I would like to invite you to our next fortnightly Open NEST Developer
Video Conference, today
Monday 1 March, 11.30-12.30 CET (UTC+1).
As usual, in the Project team round, a contact person of each team will
give a short statement summarizing ongoing work in the team and
cross-cutting points that need discussion among the teams. The remainder
of the meeting we would go into a more in-depth discussion of topics
that came up on the mailing list or that are suggested by the teams.
Today this would for example be a discussion about changing the minimum
required Python version.
Agenda
Welcome
Review of NEST User Mailing List
Project team round
In-depth discussion
Raise minimal Python version required: hike to 3.7 or 3.8 (see
#1932
<https://github.com/nest/nest-simulator/pull/1932#issuecomment-782710952>)?
The agenda for this meeting is also available online, see
https://github.com/nest/nest-simulator/wiki/2021-03-01-Open-NEST-Developer-…
Looking forward to seeing you soon!
best,
Dennis Terhorst
------------------
Log-in information
------------------
We use a virtual conference room provided by DFN (Deutsches Forschungsnetz).
You can use the web client to connect. We however encourage everyone to
use a headset for better audio quality or even a proper video
conferencing system (see below) or software when available.
Web client
* Visit https://conf.dfn.de/webapp/conference/97938800
* Enter your name and allow your browser to use camera and microphone
* The conference does not need a PIN to join, just click join and you're in.
In case you see a dfnconf logo and the phrase "Auf den
Meetingveranstalter warten", just be patient, the meeting host needs to
join first (a voice will tell you).
VC system/software
How to log in with a video conferencing system, depends on you VC system
or software.
- Using the H.323 protocol (eg Polycom): vc.dfn.net##97938800 or
194.95.240.2##97938800
- Using the SIP protocol:97938800@vc.dfn.de
- By telephone: +49-30-200-97938800
For those who do not have a video conference system or suitable
software, Polycom provides a pretty good free app for iOS and Android,
so you can join from your tablet (Polycom RealPresence Mobile, available
from AppStore/PlayStore). Note that firewalls may interfere with
videoconferencing in various and sometimes confusing ways.
For more technical information on logging in from various VC systems,
please see
http://vcc.zih.tu-dresden.de/index.php?linkid=1.1.3.4
Hello,
my name is Arturo and I think I found a bug when calling PlotTargets using
the mask plot option with an azimuthal angle > 0.
Below I give an example very easy to follow that shows that although the
targets are plotted correctly, the mask is not.
Correct me if I am doing anything wrong please. I attach an image of the
results also.
Thank you, Arturo.
Code:
l = topo.CreateLayer({'rows': 21, 'columns': 21,
'elements': 'iaf_psc_alpha'})
conndict = {'connection_type': 'divergent',
'mask': {'rectangular': {'lower_left': [-0.3, -0.12],
'upper_right': [-0.05, 0.12],
'azimuth_angle': 45.}},
'kernel': 1.0}
topo.ConnectLayers(l, l, conndict)
fig = topo.PlotLayer(l, nodesize=40)
ctr = topo.FindCenterElement(l)
topo.PlotTargets(ctr, l, fig=fig,
mask=conndict['mask'],
src_size=250, tgt_color='red', tgt_size=20)
[image: Captura.PNG]
Dear NEST Developers!
Thanks to Lekshmi's great work, we can now use Github checks for continuous integration testing. Please merge master into your open PRs/branches to make sure your contributions run the Github checks. We will run Github checks and Travis checks in parallel for now.
Best
Hans Ekkehard
--
Prof. Dr. Hans Ekkehard Plesser
Head, Department of Data Science
Faculty of Science and Technology
Norwegian University of Life Sciences
PO Box 5003, 1432 Aas, Norway
Phone +47 6723 1560
Email hans.ekkehard.plesser(a)nmbu.no<mailto:hans.ekkehard.plesser@nmbu.no>
Home http://arken.nmbu.no/~plesser
Hi!
Just a question, I may have ovelooked something. If I use NEST help to get information about a neuron model, I get the message that there is no help:
In [28]: nest.help('iaf_psc_alpha')
Sorry, there is no help for 'iaf_psc_alpha'.
But we have all the nice model documentation. Have I overlooked something or do we need to upgrade help()?
Best,
Hans Ekkehard
--
Prof. Dr. Hans Ekkehard Plesser
Head, Department of Data Science
Faculty of Science and Technology
Norwegian University of Life Sciences
PO Box 5003, 1432 Aas, Norway
Phone +47 6723 1560
Email hans.ekkehard.plesser(a)nmbu.no<mailto:hans.ekkehard.plesser@nmbu.no>
Home http://arken.nmbu.no/~plesser
Dear Johan,
You do not tell us what the “result” number is that you print out. But I notice that the value you get when running with 10 MPI processes on PDC, is about 10 times smaller than the value you see with one process on your PC.
I’ll venture a guess and assume you have a network of N neurons which you connect to one spike detector. You simulate, read out the number of spikes and convert it into a firing rate, dividing number of spikes by N * simtime.
In an MPI-parallel simulation, the spike detector on each MPI process only gets the spikes of neurons simulated on that MPI process. If you still divide by the full number of neurons to compute the rate, one would expect exactly the behavior you observe.
In the future, please provide more details about what you are doing, so we can provide more pointed advice.
Best,
Hans Ekkehard
--
Prof. Dr. Hans Ekkehard Plesser
Head, Department of Data Science
Faculty of Science and Technology
Norwegian University of Life Sciences
PO Box 5003, 1432 Aas, Norway
Phone +47 6723 1560
Email hans.ekkehard.plesser(a)nmbu.no<mailto:hans.ekkehard.plesser@nmbu.no>
Home http://arken.nmbu.no/~plesser
On 19/02/2021, 22:20, "JOHAN LILJEFORS" <JOHAN(a)liljefors.eu<mailto:JOHAN@liljefors.eu>> wrote:
Dear Nest users,
I am fairly new to Nest having used it only for a few months and it has been working without any problems on my PC running Linux. Recently I was given access to a PDC where I have been submitting jobs but I am encountering a problem with my model. I'm running Python 3.7.3 and Nest 2.18.
On both my PC and on the PDC, I run the following python script:
------------------------------
import Nest
For counter in range(0,10):
nest.resetkernel()
___run___script
print(result)
------------------------------
I execute this with "srun -n=1 python3 run.py"
This outputs 10 numbers, ranging from 0.35-0.4.
On the PDC, i run the same scripts but without the for loop:
------------------------------
import Nest
nest.resetkernel()
___run___script
print(result)
------------------------------
I execute this with "srun -n=10 python3 run.py"
this outputs 10 numbers, but this time much smaller around 0.04.
I am not doing any file operations, nor any MPI communication between the tasks and I am genuinely confused as to how submitting multiple tasks can yield a different result than a single task.
Has anyone encountered anything similar?
Regards
Johan Liljefors