|
|
Dear Danila,
the macro particles reported by `PIConGPOVerbose` is only a estimation and not take care of the gas profile. This estimation assume that each species is placed has homogeneous gas over the full simulation volume. We use it as a quick check that we not have to many macro particles in our simulation. Rule of thumb is here 10*10<6 macro particles per 1GiB gpu RAM.
The hdf5 dump (not checkpoint) contains only **macro** particles and fields within the virtual window. Each particle in the hdf5 file has it's own weighting. The weighting defines the number of real particles for this macro particle.
The is also a plugin avail which counts the macro particles on all gpus (also outside the virtual window) [1]
best reguards,
René
[1] https://github.com/ComputationalRadiationPhysics/picongpu/blob/release-0.1.2/doc/TBG_macros.cfg#L157-L160
On 11/01/2016 11:29 AM, Khikhlukha Danila wrote:
Dear René,
Thanks for your prompt replay.
So if I got your answer correctly, then the number of macro particles reported by picongpu in a PIConGPUVerbose (PHYSICS) section and number of particles in hdf5 dump (let’s say in openPMD standard) are different ( PIConGPUVerbose shows the number of particles with respect to the full domain). What is happening to macro particle's weight then? Is it being scaled?
Thanks,
Danila.
-----Original Message-----
From: picongpu-users@hzdr.de [mailto:picongpu-users@hzdr.de] On Behalf Of Rene Widera
Sent: Tuesday, November 01, 2016 11:16 AM
To: picongpu-users@hzdr.de
Subject: Re: [PIConGPU-Users] [PIConGPU-Users] Moving window notation
Dear Danila,
the domains which are reserved for the moving window are fully part of the simulation and contains fields and particles.
The most plugins currently not take care that only n-1 gpus
(VirtualWindow) are analyzed in y direction. There is an open issue [1] for that.
The hdf5 and adios output take care of the reduction in y direction and dumps out only the areas inside VirtualWindow.
best regards,
René
[1] https://github.com/ComputationalRadiationPhysics/picongpu/issues/89
On 11/01/2016 10:59 AM, Khikhlukha Danila wrote:
Dear all,
The moving window concept is described on this page:
https://github.com/ComputationalRadiationPhysics/picongpu/wiki/PIConGPU-domain-definitions#moving-window.
It is said there that the y-size of the domain if reduced by 1 GPU.
Am I correct that there are no macro particles in that area as it is
reserved for the movement computation?
Best regards,
Danila.
--
René Widera
Abteilung Laser-Teilchenbeschleunigung (FWKT) Helmholtz-Zentrum Dresden-Rossendorf
Tel: +49 (0351) 260 3543
r.widera@hzdr.de
http://www.hzdr.de
Vorstand: Prof. Dr. Dr. h. c. Roland Sauerbrey,
Prof. Dr. Dr. h. c. Peter Joehnk
Vereinsregister: VR 1693 beim Amtsgericht Dresden
#############################################################
This message is sent to you because you are subscribed to
the mailing list <picongpu-users@hzdr.de>.
To unsubscribe, E-mail to: <picongpu-users-off@hzdr.de> To switch to the DIGEST mode, E-mail to <picongpu-users-digest@hzdr.de> To switch to the INDEX mode, E-mail to <picongpu-users-index@hzdr.de> Send administrative queries to <picongpu-users-request@hzdr.de>
#############################################################
This message is sent to you because you are subscribed to
the mailing list <picongpu-users@hzdr.de>.
To unsubscribe, E-mail to: <picongpu-users-off@hzdr.de>
To switch to the DIGEST mode, E-mail to <picongpu-users-digest@hzdr.de>
To switch to the INDEX mode, E-mail to <picongpu-users-index@hzdr.de>
Send administrative queries to <picongpu-users-request@hzdr.de>
--
René Widera
Abteilung Laser-Teilchenbeschleunigung (FWKT)
Helmholtz-Zentrum Dresden-Rossendorf
Tel: +49 (0351) 260 3543
r.widera@hzdr.de
http://www.hzdr.de
Vorstand: Prof. Dr. Dr. h. c. Roland Sauerbrey,
Prof. Dr. Dr. h. c. Peter Joehnk
Vereinsregister: VR 1693 beim Amtsgericht Dresden
|
|