January 2014 - Allocations-admins - mailman3.cels.anl.gov

Re: [allocations-admins] [LCRC Accounts] Project Allocation Request
by John Valdes 31 Jan '14

31 Jan '14

FYI, Gary was asking for more time for the current quarter for project dynamics, as he's run out, so I submitted this request on his behalf based on his burn rate so far this FY. I'll grant this request for 90K shortly. John On Fri, Jan 31, 2014 at 04:43:41PM -0600, accounts(a)lcrc.anl.gov wrote: > Hello, > > A change in allocation has been requested: > > Requester: leaf (Gary Leaf) > Project: dynamics > Title: Dynamics of biophysical systems. > Description: The goal of our computations for the reduced model of a sodium channel is to compute the dependence of the free energy components for the channel. We investigate the dependence on such factors as channel size, cation size in relation to channel size, ion charge, protein polarization effects, and polarization effects in the channel. The object is to gain a deeper understanding of the selectivity mechanism whereby some ions are selected over others for transport through the channel. This is in spite of the difference in concentration outside the channel. > Our computations are based on using a Grand Canonical Monte Carlo (GCMC) scheme to compute the free energy components as well as the expected ionic concentration in the filter portion of the channel. For a given state (configuration and parameter settings) we compute a multiple set of final realizations in parallel. The number of realizations ranges from 24 to 48 with 32 being the usual number. We expect each GCMC calculation will require on the order of 40K steps with 50-70 particle insertion/deletion attempts as part of the free energy calculation. > Our computational study will divided into three main segments. Each segment has three parts that are essentially the same for each segment. > > Segment I. Here we model with a homogeneous dielectric; so that there is no polarization effects from either the protein or the or the channel. The object is to isolate the effect of excluded volume on selectivity. Because this model is homogeneous, the computations are relatively short. On the order of 8 hours per calculation. > Part A. Vary the filter pore radius: R = 3.0 3.5 4.0 4.5 5.0 Angs > Part B. Vary the type of cation competing with sodium. > Part C. Study the transmission of sodium over calcium for a sodium channel by gradually titrating calcium into the system. > Because the this model homogeneous, the computations are very modest. We estimate that the study for segment I will require on the order > of 4000 core-hours using 32 cores per computation. > Cost: 4000 core-hours for Segment I study. > > Segment II. Study the role of protein polarization by varying the protein dielectric. This model is inhomogeneous, hence considerably more computationally expensive. This study has the same three parts as Segment I, but more calculations for each part as well as the increased cost of each run. We estimate that this portion of the study will involve about 60 cases covering the parameter ranges. Using 32 cores, although 48 may be needed, each will require on the order of 1664-2496 core-hours. Based on 60 cases this yields > Cost : 99840-149760 core-hours for Segment II study > > Segment III. Study the role of polarization in the sodium channel by varying the channel dielectric. These cases also involve an inhomogeneous model; thus the cost will be > comparable to the cost of the study in Segment II. > Cost : 99840-149760 core-hours for Segment III study. > > > Current: undetermined amount > Justification: > > Requested: 90000 > > A specific reason has been given: > Our allocation for the 1st half of the year is nearly exhausted. We've used approximately 45K per month on average so far this FY; continuing at that rate, we will require another 90K for the remainder of the current quarter. > > This needs to be approved and the final allocation amount decided upon. > > Thank You, > The LCRC Accounts System > _______________________________________________ > allocations-admins mailing list > allocations-admins(a)lists.lcrc.anl.gov > https://lists.lcrc.anl.gov/mailman/listinfo/allocations-admins

1 0

[LCRC Accounts] Project Allocation Request
by accounts＠lcrc.anl.gov 31 Jan '14

31 Jan '14

Hello, A change in allocation has been requested: Requester: leaf (Gary Leaf) Project: dynamics Title: Dynamics of biophysical systems. Description: The goal of our computations for the reduced model of a sodium channel is to compute the dependence of the free energy components for the channel. We investigate the dependence on such factors as channel size, cation size in relation to channel size, ion charge, protein polarization effects, and polarization effects in the channel. The object is to gain a deeper understanding of the selectivity mechanism whereby some ions are selected over others for transport through the channel. This is in spite of the difference in concentration outside the channel. Our computations are based on using a Grand Canonical Monte Carlo (GCMC) scheme to compute the free energy components as well as the expected ionic concentration in the filter portion of the channel. For a given state (configuration and parameter settings) we compute a multiple set of final realizations in parallel. The number of realizations ranges from 24 to 48 with 32 being the usual number. We expect each GCMC calculation will require on the order of 40K steps with 50-70 particle insertion/deletion attempts as part of the free energy calculation. Our computational study will divided into three main segments. Each segment has three parts that are essentially the same for each segment. Segment I. Here we model with a homogeneous dielectric; so that there is no polarization effects from either the protein or the or the channel. The object is to isolate the effect of excluded volume on selectivity. Because this model is homogeneous, the computations are relatively short. On the order of 8 hours per calculation. Part A. Vary the filter pore radius: R = 3.0 3.5 4.0 4.5 5.0 Angs Part B. Vary the type of cation competing with sodium. Part C. Study the transmission of sodium over calcium for a sodium channel by gradually titrating calcium into the system. Because the this model homogeneous, the computations are very modest. We estimate that the study for segment I will require on the order of 4000 core-hours using 32 cores per computation. Cost: 4000 core-hours for Segment I study. Segment II. Study the role of protein polarization by varying the protein dielectric. This model is inhomogeneous, hence considerably more computationally expensive. This study has the same three parts as Segment I, but more calculations for each part as well as the increased cost of each run. We estimate that this portion of the study will involve about 60 cases covering the parameter ranges. Using 32 cores, although 48 may be needed, each will require on the order of 1664-2496 core-hours. Based on 60 cases this yields Cost : 99840-149760 core-hours for Segment II study Segment III. Study the role of polarization in the sodium channel by varying the channel dielectric. These cases also involve an inhomogeneous model; thus the cost will be comparable to the cost of the study in Segment II. Cost : 99840-149760 core-hours for Segment III study. Current: undetermined amount Justification: Requested: 90000 A specific reason has been given: Our allocation for the 1st half of the year is nearly exhausted. We've used approximately 45K per month on average so far this FY; continuing at that rate, we will require another 90K for the remainder of the current quarter. This needs to be approved and the final allocation amount decided upon. Thank You, The LCRC Accounts System

1 0

Re: [allocations-admins] [LCRC Accounts] Project Allocation Request
by John Valdes 16 Jan '14

16 Jan '14

Granted 150K for Q2. John On Tue, Jan 14, 2014 at 08:25:22AM -0600, accounts(a)lcrc.anl.gov wrote: > > A change in allocation has been requested: > > Requester: aithal (Shashi Aithal) > Project: PETSc-Foam [...] > Requested: 150000

1 0

[LCRC Accounts] Project Allocation Request
by accounts＠lcrc.anl.gov 14 Jan '14

14 Jan '14

Hello, A change in allocation has been requested: Requester: aithal (Shashi Aithal) Project: PETSc-Foam Title: Investigation of robust, scalable solvers for OpenFoam Description: OF is parallel, finite-volume CFD code. OF version 2.1.0 has been installed and tested on Fusion. Preliminary scaling studies on dsmcFoam and dieselFoam have shown good scalability. Small DSMC cases (1 million cells) have shown good scalability up to 192 cores. DieselFoam has shown good scalability up to 512 cores with about 500K cells. We intend to use OpenFoam 2.2-x as the starting point in our work. Various algorithms for efficient domain decomposition and solution methods for combusting mixtures will be investigated for optimal performance and robustness. The first quarter will focus on compiling, running and testing small problems and studying various methodologies. More detailed cases studies are planned for the second, third and fourth quarters. Current: undetermined amount Justification: Requested: 150000 A specific reason has been given: The current allocation of 150K core-hours was used to refine a Sigma LES model in OpenFoam. A typical case needs about 3000 core hours and about 40 such cases have been run. Good validation results of ERCOFTAC cases have been obtained. The additional time is being requested to apply the LES model to actual engine simulations and evaluate the computational effort. A typical engine case is expected to take about 30K core-hours and 5 such runs are planned to complete the initial evaluation of the LES model for actual engine geometries. This needs to be approved and the final allocation amount decided upon. Thank You, The LCRC Accounts System

1 0

Re: [allocations-admins] [LCRC Accounts] Project Allocation Request
by Bair, Raymond A. 02 Jan '14

02 Jan '14

Steve saysŠ From: Steve Kuhlmann <steve.kuhlmann1(a)gmail.com> Date: Thursday, January 2, 2014 3:00 PM To: Ray Bair <rbair(a)anl.gov> Subject: Re: [allocations-admins] [LCRC Accounts] Project Allocation Request Hi Ray. I think we are far from having an optimized queuing, but one successful sequence was 48 nodes/384 cpu, and we had 3 jobs going at the same time with this allocation. Steve Ray asksŠ This project was granted 83K hours per quarter (332 total) for FY14. It appears that they want to add 200K for the 2nd half. Should we add 100K now for Q3 and then take the rest to the Committee in March? Ray > > > > > > >On 1/2/14 9:21 AM, "accounts(a)lcrc.anl.gov" <accounts(a)lcrc.anl.gov> wrote: > >>Hello, >> >>A change in allocation has been requested: >> >> Requester: kuhlmann (Stephen Kuhlmann) >> Project: Supernova_DES >> Title: Supernova modeling and analysis for Dark Energy Survey >> Description: The project is to define the supernova >>cosmology uncertainties for the new Dark Energy Survey which will >>collect >>thousands of >>supernova explosions. >> >>This project uses the SNANA simulation package >>http://sdssdp62.fnal.gov/sdsssn/SNANA-PUBLIC/ >>as well as the SNCosmo analysis package >>http://sncosmo.readthedocs.org/en/latest/index.html. >>Both packages have been compiled on Fusion. >> >>The light-curve fits in SNANA, to a single >>supernova, takes 30s on a dedicated processor. >>We need x10 the expected data statistics in our >>simulations to define uncertainties. >>The total expected simulation load per quarter is: >>50K supernova * 30s * 10 parameters * 20 sets = >>83K hours. Each quarter our simulation >>will be updated due to improved >>knowledge based on the new data now being >>recorded. >> >>This project will involve five members of the >>ANL/HEP DES collaboration. >> >> >> >> Current: undetermined amount >>Justification: >> >> Requested: 200000 >> >>A specific reason has been given: >>The previous 166K hours on fusion have been well used. The results form >>the basis of a presentation at the most >>prestigious astronomy conference next >>week in Washington DC. In addition, we upgraded our software to use >>MPI >>and most of the previous allocation was >>used efficiently. We have learned >>a lot from the previous results and are >>preparing a new set of supernova >>simulations and analysis that require a >>new allocation of comparable size. >> >>This needs to be approved and the final allocation amount decided upon. >> >>Thank You, >>The LCRC Accounts System >>_______________________________________________ >>allocations-admins mailing list >>allocations-admins(a)lists.lcrc.anl.gov >>https://lists.lcrc.anl.gov/mailman/listinfo/allocations-admins

1 0

Re: [allocations-admins] [LCRC Accounts] Project Allocation Request
by Bair, Raymond A. 02 Jan '14

02 Jan '14

FYI, I wrote to Stephen to ask about how many cores he plans on using. Ray On 1/2/14 9:21 AM, "accounts(a)lcrc.anl.gov" <accounts(a)lcrc.anl.gov> wrote: >Hello, > >A change in allocation has been requested: > > Requester: kuhlmann (Stephen Kuhlmann) > Project: Supernova_DES > Title: Supernova modeling and analysis for Dark Energy Survey > Description: The project is to define the supernova >cosmology uncertainties for the new Dark Energy Survey which will collect >thousands of >supernova explosions. > >This project uses the SNANA simulation package >http://sdssdp62.fnal.gov/sdsssn/SNANA-PUBLIC/ >as well as the SNCosmo analysis package >http://sncosmo.readthedocs.org/en/latest/index.html. >Both packages have been compiled on Fusion. > >The light-curve fits in SNANA, to a single >supernova, takes 30s on a dedicated processor. >We need x10 the expected data statistics in our >simulations to define uncertainties. >The total expected simulation load per quarter is: >50K supernova * 30s * 10 parameters * 20 sets = >83K hours. Each quarter our simulation >will be updated due to improved >knowledge based on the new data now being >recorded. > >This project will involve five members of the >ANL/HEP DES collaboration. > > > > Current: undetermined amount >Justification: > > Requested: 200000 > >A specific reason has been given: >The previous 166K hours on fusion have been well used. The results form >the basis of a presentation at the most >prestigious astronomy conference next >week in Washington DC. In addition, we upgraded our software to use MPI >and most of the previous allocation was >used efficiently. We have learned >a lot from the previous results and are >preparing a new set of supernova >simulations and analysis that require a >new allocation of comparable size. > >This needs to be approved and the final allocation amount decided upon. > >Thank You, >The LCRC Accounts System >_______________________________________________ >allocations-admins mailing list >allocations-admins(a)lists.lcrc.anl.gov >https://lists.lcrc.anl.gov/mailman/listinfo/allocations-admins

1 0

[LCRC Accounts] Project Allocation Request
by accounts＠lcrc.anl.gov 02 Jan '14

02 Jan '14

Hello, A change in allocation has been requested: Requester: kuhlmann (Stephen Kuhlmann) Project: Supernova_DES Title: Supernova modeling and analysis for Dark Energy Survey Description: The project is to define the supernova cosmology uncertainties for the new Dark Energy Survey which will collect thousands of supernova explosions. This project uses the SNANA simulation package http://sdssdp62.fnal.gov/sdsssn/SNANA-PUBLIC/ as well as the SNCosmo analysis package http://sncosmo.readthedocs.org/en/latest/index.html. Both packages have been compiled on Fusion. The light-curve fits in SNANA, to a single supernova, takes 30s on a dedicated processor. We need x10 the expected data statistics in our simulations to define uncertainties. The total expected simulation load per quarter is: 50K supernova * 30s * 10 parameters * 20 sets = 83K hours. Each quarter our simulation will be updated due to improved knowledge based on the new data now being recorded. This project will involve five members of the ANL/HEP DES collaboration. Current: undetermined amount Justification: Requested: 200000 A specific reason has been given: The previous 166K hours on fusion have been well used. The results form the basis of a presentation at the most prestigious astronomy conference next week in Washington DC. In addition, we upgraded our software to use MPI and most of the previous allocation was used efficiently. We have learned a lot from the previous results and are preparing a new set of supernova simulations and analysis that require a new allocation of comparable size. This needs to be approved and the final allocation amount decided upon. Thank You, The LCRC Accounts System

1 0