Marco Baldi & Kazuya Koyama CSWG Task-6: Numerical tools for - - PowerPoint PPT Presentation
Marco Baldi & Kazuya Koyama CSWG Task-6: Numerical tools for non-standard cosmological scenarios Second report on the activities of the CSWG-T6 Euclid CSWG-OUSIM joint kick-off meeting Barcelona, 9 VII 2012 aims of the CSWG-T6 Marco
Second report on the activities
Euclid CSWG-OUSIM joint kick-off meeting Barcelona, 9 VII 2012
Marco Baldi - Euclid Collaboration Meeting, Copenhagen, 16 V 2012
1) Develop, test, and optimize numerical tools (N-body codes, Initial Conditions generators, Halo finders and other post-processing algorithms) suitable to investigate cosmological scenarios beyond the plain standard ΛCDM as defined by the Theory WG
Marco Baldi - Euclid Collaboration Meeting, Copenhagen, 16 V 2012
1) Develop, test, and optimize numerical tools (N-body codes, Initial Conditions generators, Halo finders and other post-processing algorithms) suitable to investigate cosmological scenarios beyond the plain standard ΛCDM as defined by the Theory WG 2) Coordinate the long-term simulations program for such models by defining the appropriate numerical parameters for different types of cosmological simulations and a standardized format for the output deliverables, in close collaboration with the other Tasks of the CSWG
Marco Baldi - Euclid Collaboration Meeting, Copenhagen, 16 V 2012
1) Develop, test, and optimize numerical tools (N-body codes, Initial Conditions generators, Halo finders and other post-processing algorithms) suitable to investigate cosmological scenarios beyond the plain standard ΛCDM as defined by the Theory WG 2) Coordinate the long-term simulations program for such models by defining the appropriate numerical parameters for different types of cosmological simulations and a standardized format for the output deliverables, in close collaboration with the other Tasks of the CSWG 3) Coordinate science applications of the simulations of non-standard cosmological models in connection with all the different SWGs
Marco Baldi - Euclid Collaboration Meeting, Copenhagen, 16 V 2012
1) Develop, test, and optimize numerical tools (N-body codes, Initial Conditions generators, Halo finders and other post-processing algorithms) suitable to investigate cosmological scenarios beyond the plain standard ΛCDM as defined by the Theory WG 2) Coordinate the long-term simulations program for such models by defining the appropriate numerical parameters for different types of cosmological simulations and a standardized format for the output deliverables, in close collaboration with the other Tasks of the CSWG 3) Coordinate science applications of the simulations of non-standard cosmological models in connection with all the different SWGs
Marco Baldi - Euclid Collaboration Meeting, Copenhagen, 16 V 2012
Presently about 15 participants
including Modified Gravity theories and massive coupled scalar field Dark Energy scenarios.
including Local, Equilateral, and Orthogonal types of non-Gaussianity in the primordial density field.
including models with a cosmological fraction of massive neutrinos.
including microscopic models of Dark Matter scattering and short-range scalar forces (Loeb & Weiner 2011).
including Coupled Dark Energy, Extended Quintessence, and Clustering Dark Energy models.
including dynamic scalar field models and phenomenological parametrizations
including LTB cosmologies parametrized by the structural properties of the density void, in particular of its radius r0.
including models with a thermal relic of particles with mass in the range mWDM=[0.5,2.0] keV.
Marco Baldi - Euclid Collaboration Meeting, Copenhagen, 16 V 2012
Marco Baldi - Euclid Collaboration Meeting, Copenhagen, 16 V 2012
1) Carry out a preliminary survey among the Task members to gather information on already existing tools and presently available simulations for each of the 8 classes of non-standard cosmologies;
Marco Baldi - Euclid Collaboration Meeting, Copenhagen, 16 V 2012
Marco Baldi - Euclid Collaboration Meeting, Copenhagen, 16 V 2012
including dynamic scalar field models and phenomenological parametrizations 5 groups actively working on class 1, some very large simulations (DEUSS) already available
gadget - based N-body codes: Baldi (Munich) / Dolag (Munich) / Carlesi & Knebe (Madrid) ramses - based N-body codes: Alimi & Corasaniti (Paris) grafic - based zoom IC generator: Penzo & Maccioʼ (Heidelberg)
including LTB cosmologies parametrized by the structural properties of the density void, in particular of its radius r0. 1 group actively working on class 2, some large simulations already available
gadget - based N-body codes: Garcia-Bellido (Madrid)
including models with a thermal relic of particles with mass in the range mWDM=[0.5,2.0] keV. 1 group actively working on class 3, some medium-size simulations already available
gadget - based N-body codes: Viel & Baldi (Trieste, Munich)
including Local, Equilateral, and Orthogonal types of non-Gaussianity in the primordial density field. 2 groups actively working on class 4, some large simulations already available
gadget - based N-body codes: Dolag (Munich) / Baldi (Munich) cubep3m: Iliev (Sussex)
Marco Baldi - Euclid Collaboration Meeting, Copenhagen, 16 V 2012
including models with a cosmological fraction of massive neutrinos. 2 groups actively working on class 5, some medium-size simulations already available
gadget - based N-body codes: Viel & Springel (Trieste, Heidelberg) / Hannestad (Aarhus)
including microscopic models of Dark Matter scattering and short-range scalar forces (Loeb & Weiner 2011). 1 group actively working on class 6, no suitable simulations available yet
N-body code: Pearce (Nottingham)
including Coupled Dark Energy, Extended Quintessence, and Clustering Dark Energy models. 1 group actively working on class 7, some large simulations (CoDECS) already available
gadget - based N-body codes: Baldi (Munich)
including Modified Gravity theories and massive coupled scalar field Dark Energy scenarios. 3 groups actively working on class 8, some small simulations already available
gadget - based N-body codes: mg-gadget Puchwein, Baldi & Springel (Heidelberg, Munich) ramses - based N-body codes: ecosmog Li, Koyama & Zhao (Durham, Portsmouth),
Llinares, Winther & Mota (Oslo)
ramses - based N-body codes: Li (Durham)
1) Carry out a preliminary survey among the Task members to gather information on already existing tools and presently available simulations for each of the 8 classes of non-standard cosmologies;
Marco Baldi - Euclid Collaboration Meeting, Copenhagen, 16 V 2012
1) Carry out a preliminary survey among the Task members to gather information on already existing tools and presently available simulations for each of the 8 classes of non-standard cosmologies;
Marco Baldi - Euclid Collaboration Meeting, Copenhagen, 16 V 2012
1) Carry out a preliminary survey among the Task members to gather information on already existing tools and presently available simulations for each of the 8 classes of non-standard cosmologies; 2) Promote and coordinate specific code comparison initiatives for those scenarios for which N-body codes have been only recently developed and are not yet sufficiently tested (as e.g. Modified Gravity codes, Clustering Dark Energy);
Marco Baldi - Euclid Collaboration Meeting, Copenhagen, 16 V 2012
Puchwein et al. 2012, in prep.
1) Carry out a preliminary survey among the Task members to gather information on already existing tools and presently available simulations for each of the 8 classes of non-standard cosmologies; 2) Promote and coordinate specific code comparison initiatives for those scenarios for which N-body codes have been only recently developed and are not yet sufficiently tested (as e.g. Modified Gravity codes, Clustering Dark Energy);
Marco Baldi - Euclid Collaboration Meeting, Copenhagen, 16 V 2012
1) Carry out a preliminary survey among the Task members to gather information on already existing tools and presently available simulations for each of the 8 classes of non-standard cosmologies; 2) Promote and coordinate specific code comparison initiatives for those scenarios for which N-body codes have been only recently developed and are not yet sufficiently tested (as e.g. Modified Gravity codes, Clustering Dark Energy);
Marco Baldi - Euclid Collaboration Meeting, Copenhagen, 16 V 2012
1) Carry out a preliminary survey among the Task members to gather information on already existing tools and presently available simulations for each of the 8 classes of non-standard cosmologies; 2) Promote and coordinate specific code comparison initiatives for those scenarios for which N-body codes have been only recently developed and are not yet sufficiently tested (as e.g. Modified Gravity codes, Clustering Dark Energy); 3) Coordinate among different research groups the production runs for those scenarios for which sufficiently well-tested codes already exist (as e.g. standard scalar field Dark Energy models);
Marco Baldi - Euclid Collaboration Meeting, Copenhagen, 16 V 2012
1) Carry out a preliminary survey among the Task members to gather information on already existing tools and presently available simulations for each of the 8 classes of non-standard cosmologies; 2) Promote and coordinate specific code comparison initiatives for those scenarios for which N-body codes have been only recently developed and are not yet sufficiently tested (as e.g. Modified Gravity codes, Clustering Dark Energy); 3) Coordinate among different research groups the production runs for those scenarios for which sufficiently well-tested codes already exist (as e.g. standard scalar field Dark Energy models); 4) Deliver numerical outputs for all the different classes of alternative cosmologies according to the standard output format that will have to be defined jointly with the other tasks of the CSWG;
Marco Baldi - Euclid Collaboration Meeting, Copenhagen, 16 V 2012
1) Carry out a preliminary survey among the Task members to gather information on already existing tools and presently available simulations for each of the 8 classes of non-standard cosmologies; 2) Promote and coordinate specific code comparison initiatives for those scenarios for which N-body codes have been only recently developed and are not yet sufficiently tested (as e.g. Modified Gravity codes, Clustering Dark Energy); 3) Coordinate among different research groups the production runs for those scenarios for which sufficiently well-tested codes already exist (as e.g. standard scalar field Dark Energy models); 4) Deliver numerical outputs for all the different classes of alternative cosmologies according to the standard output format that will have to be defined jointly with the other tasks of the CSWG;
Marco Baldi - Euclid Collaboration Meeting, Copenhagen, 16 V 2012
1) Carry out a preliminary survey among the Task members to gather information on already existing tools and presently available simulations for each of the 8 classes of non-standard cosmologies; 2) Promote and coordinate specific code comparison initiatives for those scenarios for which N-body codes have been only recently developed and are not yet sufficiently tested (as e.g. Modified Gravity codes, Clustering Dark Energy); 3) Coordinate among different research groups the production runs for those scenarios for which sufficiently well-tested codes already exist (as e.g. standard scalar field Dark Energy models); 4) Deliver numerical outputs for all the different classes of alternative cosmologies according to the standard output format that will have to be defined jointly with the other tasks of the CSWG;
Marco Baldi - Euclid Collaboration Meeting, Copenhagen, 16 V 2012
Marco Baldi - Euclid Collaboration Meeting, Copenhagen, 16 V 2012
classes of models, possibly multiple specific models for each class, sampling the parameter space for each model; this could easily sum up to several tens of large N-body runs: suitable computational resources should be allocated for this task, both at the level of the Euclid Consortium and at the level of the member countries;
Marco Baldi - Euclid Collaboration Meeting, Copenhagen, 16 V 2012
classes of models, possibly multiple specific models for each class, sampling the parameter space for each model; this could easily sum up to several tens of large N-body runs: suitable computational resources should be allocated for this task, both at the level of the Euclid Consortium and at the level of the member countries;
(mock catalogs, power spectra, light cones, lensing maps, etc...) for a wide range of non-standard cosmologies, which should be used for science applications by the whole Euclid collaboration: need to identify appropriate options for long-term data storage and fast data-sharing networks;
Marco Baldi - Euclid Collaboration Meeting, Copenhagen, 16 V 2012
classes of models, possibly multiple specific models for each class, sampling the parameter space for each model; this could easily sum up to several tens of large N-body runs: suitable computational resources should be allocated for this task, both at the level of the Euclid Consortium and at the level of the member countries;
(mock catalogs, power spectra, light cones, lensing maps, etc...) for a wide range of non-standard cosmologies, which should be used for science applications by the whole Euclid collaboration: need to identify appropriate options for long-term data storage and fast data-sharing networks;
Marco Baldi - Euclid Collaboration Meeting, Copenhagen, 16 V 2012
the CSWG, since all the numerical developments (code optimization, halo finders, light-cone construction, accurate power spectrum computation, etc...) for ΛCDM runs will have to be applied also to non- standard simulations;
classes of models, possibly multiple specific models for each class, sampling the parameter space for each model; this could easily sum up to several tens of large N-body runs: suitable computational resources should be allocated for this task, both at the level of the Euclid Consortium and at the level of the member countries;
Marco Baldi - Euclid Collaboration Meeting, Copenhagen, 16 V 2012
classes of models, possibly multiple specific models for each class, sampling the parameter space for each model; this could easily sum up to several tens of large N-body runs: suitable computational resources should be allocated for this task, both at the level of the Euclid Consortium and at the level of the member countries;
Marco Baldi - Euclid Collaboration Meeting, Copenhagen, 16 V 2012
Some groups have already started to allocate dedicated resources:
classes of models, possibly multiple specific models for each class, sampling the parameter space for each model; this could easily sum up to several tens of large N-body runs: suitable computational resources should be allocated for this task, both at the level of the Euclid Consortium and at the level of the member countries;
Marco Baldi - Euclid Collaboration Meeting, Copenhagen, 16 V 2012
Some groups have already started to allocate dedicated resources: The INCAS network (P.I. M. Baldi) has submitted a CLASS-A PRACE proposal for 25 Mio CPU hours on FERMI devoted to DE/MG simulations (gadget-3, gadget-3c, mg-gadget)
classes of models, possibly multiple specific models for each class, sampling the parameter space for each model; this could easily sum up to several tens of large N-body runs: suitable computational resources should be allocated for this task, both at the level of the Euclid Consortium and at the level of the member countries;
Marco Baldi - Euclid Collaboration Meeting, Copenhagen, 16 V 2012
Some groups have already started to allocate dedicated resources: The INCAS network (P.I. M. Baldi) has submitted a CLASS-A PRACE proposal for 25 Mio CPU hours on FERMI devoted to DE/MG simulations (gadget-3, gadget-3c, mg-gadget) The PORTSMOUTH group (news from G.B. Zhao) has access to an in-house supercomputing system (the SCIAMA cluster, 1008 cores, 2Gb/core) largely devoted to MG simulations (ecosmog, mlapm)
CoDECS
Marco Baldi - Euclid Collaboration Meeting, Copenhagen, 16 V 2012
We aim for something like this for all the 8 classes
cosmological scenarios!
If you are interested in the activities of Task 6 and want to be included in our mailing list, please drop an email to the coordinators: Marco Baldi: marco.baldi@universe-cluster.de Kazuya Koyama: Kazuya.Koyama@port.ac.uk CoDECS
Marco Baldi - Euclid Collaboration Meeting, Copenhagen, 16 V 2012
We aim for something like this for all the 8 classes
cosmological scenarios!