Links

Content Skeleton

This Page

Previous topic

Muon Simulation Presentation

Next topic

optix

GTC Preparation

Intro

I will introduce Opticks, an optical photon simulation based on OptiX, that enables neutrino physics to benefit from from massive parallelism

Dayabay Intro

Described as “liquid onions,” each detector will contain three layers. In the center will be 20 tons of organic liquid scintillator that contains gadolinium, a heavy metal. Next is a layer of liquid scintillator without gadolinium. The outer layer uses mineral oil to act as shielding. When an antineutrino interacts with an atom inside the detector, it produces a positron and a neutron. The energy from the positron is deposited in the scintillator, which creates a burst of light. About 30 microseconds later, a second burst of light is produced as the gadolinium captures the neutron and amplifies the signal. Photomultiplier tubes that line the mineral-oil-filled outer detector tank record the light produced in this reaction. Both light flashes must be present to indicate an electron antineutrino event.

Potential Users for fast Optical Photon Simulation

GATE is an advanced opensource software developed by the international OpenGATE collaboration and dedicated to numerical simulations in medical imaging and radiotherapy. It currently supports simulations of Emission Tomography (Positron Emission Tomography - PET and Single Photon Emission Computed Tomography - SPECT), Computed Tomography (CT), Optical Imaging (Bioluminescence and Fluorescence) and Radiotherapy experiments.

Talk coaching 3

Starting : 1 min

In a very short time, establish:

  • Attention
  • Interest
  • Credibility
  • Expectations
    • how my technology works and how this can change the ... industry
    • show you how Opticks enables particle physics (especially neutrino physics) to start to really benefit from GPUs

Transitions

  • now that i have introduced ... i will move on to using ...

Concluding : 30s

Signalling the end

  • recap : what you said : very high level
  • bookends : the intro and conclusion should be similar
  • clincher statements

Abstract

Opticks : Optical Photon Simulation for High Energy Physics with OptiX

Opticks is an open source project that integrates high performance GPU
ray tracing from NVIDIA OptiX with existing Geant4 toolkit based simulations.
Advantages of separate optical photon simulation and
the approaches developed to integrate it with the general Geant4
particle simulation are presented. Approaches to minimize overheads
arising from split are shown.
Challenges included bringing complex geometries with wavelength
dependent material and surface properties to the GPU.
Techniques for visualisation of photon propagations with
interactive time scrubbing and history selection using OpenGL/OptiX/Thrust
interoperation and geometry shaders are described.
Results and demonstrations are shown for the photomultiplier based
Daya Bay and JUNO Neutrino detectors.
Extrapolation of observed timings with test geometries to multi-GPU workstation
core counts suggests performance sufficient for optical photon processing time to
become effectively zero compared to total simulation time is within reach.

Title Image

~/simon

jpmt-inside-wide.png 2864x1688

delta:doc blyth$ crop.py --style custom --width 2560 --height 1440 --top 0 --left 0  ~/simoncblyth.bitbucket.org/env/graphics/ggeoview/jpmt-inside-wide.png
2016-03-26 12:18:25,123 env.doc.crop INFO     Crop None custom
2016-03-26 12:18:25,123 env.doc.crop INFO     cropping /Users/blyth/simoncblyth.bitbucket.org/env/graphics/ggeoview/jpmt-inside-wide.png to create /Users/blyth/simoncblyth.bitbucket.org/env/graphics/ggeoview/jpmt-inside-wide_crop.png
2016-03-26 12:18:25,125 env.doc.crop INFO     width 2864 height 1688 cropping to box (0, 0, 2560, 1440)
delta:doc blyth$

target 1280x720 target*2 2560x1440

SK

The Super-Kamiokande detector is massive, even by particle physics standards. It consists of 50,000 tons of pure water surrounded by about 11,200 photomultiplier tubes. The detector was again designed as a cylindrical structure, this time 41.4 m tall and 39.3 m across.

curl -L -O http://www-sk.icrr.u-tokyo.ac.jp/sk/gallery/wme/PH20-water-withboat-apr23-wm.jpg

source 3872x2592
target 2560x1440

width  3872 - 2560 = 1312  , 1312/2 = 656
height 2592 - 1440 = 1152  , 1152/2 = 576


crop.py --style custom --width 2560 --height 1440 --top 656 --left 576 PH20-water-withboat-apr23-wm.jpg

Central crop decreases image impact:

delta:tmp blyth$ crop.py --style custom --width 2560 --height 1440 --top 656 --left 576 PH20-water-withboat-apr23-wm.jpg
2016-03-26 15:02:27,219 env.doc.crop INFO     Crop None custom
2016-03-26 15:02:27,220 env.doc.crop INFO     cropping PH20-water-withboat-apr23-wm.jpg to create PH20-water-withboat-apr23-wm_crop.jpg
2016-03-26 15:02:27,221 env.doc.crop INFO     width 3872 height 2592 cropping to box (576, 656, 3136, 2096)
delta:tmp blyth$

Try just trimming the base to give 16:9:

In [5]: 3872./2592.
Out[5]: 1.4938271604938271

In [6]: 3872./16
Out[6]: 242.0

In [7]: 3872./16*9
Out[7]: 2178.0


delta:tmp blyth$ crop.py --style custom --width 3872 --height 2178 --top 0 --left 0 PH20-water-withboat-apr23-wm.jpg
2016-03-26 15:12:30,407 env.doc.crop INFO     Crop None custom
2016-03-26 15:12:30,407 env.doc.crop INFO     cropping PH20-water-withboat-apr23-wm.jpg to create PH20-water-withboat-apr23-wm_crop.jpg
2016-03-26 15:12:30,408 env.doc.crop INFO     width 3872 height 2592 cropping to box (0, 0, 3872, 2178)

delta:tmp blyth$ crop.py --style custom --width 3872 --height 2178 --top 0 --left 0 PH20-water-withboat-apr23-wm.jpg --ext .png
2016-03-26 15:23:26,170 env.doc.crop INFO     Crop None custom
2016-03-26 15:23:26,170 env.doc.crop WARNING  converting ext from .jpg to .png
2016-03-26 15:23:26,170 env.doc.crop INFO     cropping PH20-water-withboat-apr23-wm.jpg to create PH20-water-withboat-apr23-wm_crop.png
2016-03-26 15:23:26,173 env.doc.crop INFO     width 3872 height 2592 cropping to box (0, 0, 3872, 2178)
delta:tmp blyth$

3872/2
2592/2

Census of large PMT experiments, existing and planned

Make a table of large PMT expts

  • PMT size, type, number
  • collaboration dates

MCP-PMT design * http://ndip.in2p3.fr/ndip11/AGENDA/AGENDA-by-DAY/Presentations/3Wednesday/AM/ID47-Qian.pdf

Simulation of large photomultipliers for experiments in astroparticle physics * http://arxiv.org/pdf/1001.1283v2.pdf

Amanda, IceCube, IceTop, Kamiokande, NESTOR, NEMO, Antares, MiniBooNE, Xenon, Baikal GVD Tunka, North Auger Observatory

https://www.mpi-hd.mpg.de/hfm/CosmicRay/CosmicRaySites.html

https://en.wikipedia.org/wiki/List_of_neutrino_experiments

Reactor Neutrino expts

Neutrino Telescopes

Large Water Cerenkov Detectors

  • SNO
  • SuperKamiokande/HyperKamiokande

Large Scintillator Detectors

DYB Images

Far site WP filling

delta:tmp blyth$ crop.py --style custom --ext .png --width 1600 --height 900 --left 0 --top 0 DybFar.jpg
2016-03-26 17:48:59,443 env.doc.crop INFO     Crop None custom
2016-03-26 17:48:59,443 env.doc.crop WARNING  converting ext from .jpg to .png
2016-03-26 17:48:59,443 env.doc.crop INFO     cropping DybFar.jpg to create DybFar_crop.png
2016-03-26 17:48:59,445 env.doc.crop INFO     width 1600 height 900 cropping to box (0, 0, 1600, 900)
delta:tmp blyth$
delta:tmp blyth$ open DybFar_crop.png

IMPORTANT – READ CAREFULLY: This End User License Agreement (“Agreement”) is a legal agreement between you (in your capacity as an individual and as an agent for your company, institution or other entity) (collectively, “you” or “Licensee”) and The Regents of the University of California, Department of Energy contract-operators of the Ernest Orlando Lawrence Berkeley National Laboratory (“Berkeley Lab”).

Downloading, displaying, using, or copying of the image by you or by a third party on your behalf indicates your agreement to be bound by the terms and conditions of the End User License Agreement and that you have read and agree to the Copyright Notice, Disclaimers and Usage Terms.

You agree that the image selected by you may be used for noncommercial, educational purposes only; no derivative works are allowed, photos of individuals may only be used for identifying the individual and/or their research, and attribution and copyright notice is required.

Please credit the Lawrence Berkeley National Laboratory and provide the following copyright notice: “© 2010 The Regents of the University of California, through the Lawrence Berkeley National Laboratory.” If you do not agree to these terms and conditions, do not download, display or use the image.

http://www.lbl.gov/EndUserLicenseAgreement.html

SK Images

Dear Public Affairs Dept,

I am a Physicist working at the National Taiwan University, Taipei with the
Daya Bay and JUNO Collaborations.

    http://www-sk.icrr.u-tokyo.ac.jp/sk/gallery/wme/PH13-bottom-yoko-1-wm.JPG

I would like to use some of your images, such as to above, in an upcoming
presentation at the GPU technology conference in San Jose California.

     http://www.gputechconf.com

My presentation is on an open source  Optical Photon Simulation
that I have developed using NVIDIA OptiX ray tracing framework,
the abstract of my talk is below.

A technical presentation of my work is accessible below.

    http://simoncblyth.bitbucket.org/env/presentation/opticks_gpu_optical_photon_simulation_march2016.html

To make this work accessible to a diverse audience without a background in physics
I need to provide context and motivation for optical photon simulation.  I think that
introducing the audience to the use of photomultiplier tubes in several large physics
experiments including Super Kamiokande would be an excellent way
to do this, especially due to the beautiful images you have provided at

     http://www-sk.icrr.u-tokyo.ac.jp/sk/gallery/index-e.html

Sincerely,
           Dr Simon C. Blyth



Opticks : Optical Photon Simulation for High Energy Physics with OptiX

Opticks is an open source project that integrates high performance GPU
ray tracing from NVIDIA OptiX with existing Geant4 toolkit based simulations.
Advantages of separate optical photon simulation and
the approaches developed to integrate it with the general Geant4
particle simulation are presented. Approaches to minimize overheads
arising from split are shown.
Challenges included bringing complex geometries with wavelength
dependent material and surface properties to the GPU.
Techniques for visualisation of photon propagations with
interactive time scrubbing and history selection using OpenGL/OptiX/Thrust
interoperation and geometry shaders are described.
Results and demonstrations are shown for the photomultiplier based
Daya Bay and JUNO Neutrino detectors.
Extrapolation of observed timings with test geometries to multi-GPU workstation
core counts suggests performance sufficient for optical photon processing time to
become effectively zero compared to total simulation time is within reach.
Dear Dr. Simon C. Blyth,

Thank you so much for your inquiry.
There is no problem to use our image for your purpose.

Please credit as
Kamioka Observatory, ICRR(Institute for Cosmic Ray Research), The University of Tokyo.

Best regards,
Yumiko Takenaga