Ray, I have noticed another gaussian user running hundreds of single node jobs because they did not how to use Linda. Although Paul has been around for a while and I expect that he knows how to run Gaussian in parallel, it would be good to check. John J. Low Principal Computational Science Specialist Computing, Environment and Life Sciences Building 240, 2143 9700 South Cass Avenue Argonne National Laboratory Argonne, IL 60439. 630-252-0045 www.linkedin.com/pub/john-low/15/8b0/5aa/ -----Original Message----- From: <[email protected]> on behalf of "Bair, Raymond A." <[email protected]> Reply-To: LCRC Allocations Admins <[email protected]> Date: Monday, November 23, 2015 at 2:31 PM To: LCRC Allocations Admins <[email protected]> Subject: Re: [allocations-admins] [LCRC Accounts] Project Allocation Request
Yes, it is a 2-node job, and that seems fine for a Gaussian user.
It is hard to know whether Paul's work could be run on more nodes efficiently. It does not seem to be a high priority for a 100K request.
Ray
On 11/23/15, 1:21 PM, "[email protected] on behalf of Low, John J." <[email protected] on behalf of [email protected]> wrote:
Ray,
It would be good to look at the input, output, logs and scripts, as opposed to just some PBS parameters and command and a few Gaussian keywords.
This looks like a command for two node job. Not a single node job.
I would recommend that he used 60GB of memory rather than 30GB.
If gaussian really does not scale past one or two nodes for his cases, then he should consider some other program, we have MOLPRO, Orca, GAMMES and NWCHEM. By the way, what is the status of Qchem?
John J. Low Principal Computational Science Specialist Computing, Environment and Life Sciences Building 240, 2143 9700 South Cass Avenue Argonne National Laboratory Argonne, IL 60439. 630-252-0045 www.linkedin.com/pub/john-low/15/8b0/5aa/
-----Original Message----- From: <[email protected]> on behalf of "Bair, Raymond A." <[email protected]> Reply-To: LCRC Allocations Admins <[email protected]> Date: Monday, November 23, 2015 at 1:12 PM To: LCRC Allocations Admins <[email protected]> Subject: Re: [allocations-admins] [LCRC Accounts] Project Allocation Request
I queried Paul ...
Ray, My pbs file was set up as follows on Blues: #PBS -l nodes=2:ppn=16
My Gaussian jobs request memory and shared processors as follows on Blues: %mem=30GB %Nprocshared=16
Thanks, Paul
Overall it seems to be a reasonable request for 100K hours.
Ray
----------------------------------------- Ray Bair Computing, Environment, and Life Sciences Argonne National Laboratory and the University of Chicago TCS Building 240, Room 4122 9700 South Cass Avenue Argonne, IL 60439 email: rbair(at)anl.gov Phone: (630)252-5751
On 11/20/15, 5:59 PM, "[email protected] on behalf of Low, John J." <[email protected] on behalf of [email protected]> wrote:
It does not sound correct that high level ab initio methods only scale to 16 processors. Unless he means 16 16-core processors.
John J. Low Principal Computational Science Specialist Computing, Environment and Life Sciences Building 240, 2143 9700 South Cass Avenue Argonne National Laboratory Argonne, IL 60439. 630-252-0045 www.linkedin.com/pub/john-low/15/8b0/5aa/
________________________________________ From: [email protected] [[email protected]] on behalf of Aithal, Shashikant M. [[email protected]] Sent: Friday, November 20, 2015 4:06 PM To: LCRC Allocations Admins Subject: Re: [allocations-admins] [LCRC Accounts] Project Allocation Request
I think he should also submit a list of anticipated runs to be made along with the approximate time for each of those runs to justify the 100K request. It is not clear from the request how he plans to use the 100K between now and Dec 31st (first quarter) and the second quarter. His proposal also says that "...the high-level ab initio methods scale only up to 16 processors. Therefore these calculations do require a longer execution time." which means he will be running a lot of single-node jobs for a long time - which won't be good for the queue. Since he will be using only 384 core-hours/day (16x24) he would have to submit some 150 such day-long node jobs to use up 50K (half the request) between now and Dec 31st (40 days or so). He can do that if he has about 4 nodes reserved for himself till the end of Dec. ________________________________________ From: [email protected] [[email protected]] on behalf of John Blaas [[email protected]] Sent: Friday, November 20, 2015 3:44 PM To: LCRC Allocations Admins Subject: Re: [allocations-admins] [LCRC Accounts] Project Allocation Request
From Paul.
John, A lot more Gaussian 09 calculations of oxidation and reduction potentials as well as transition states for decomposition and polymerization of redoxshuttles were required than we originally anticipated.
Paul
On Fri, Nov 20, 2015 at 1:51 PM, John Blaas <[email protected]> wrote:
What you don't call this a justification?
A specific reason has been given: ran out of time
I'll ping Paul.
- JB
On Fri, Nov 20, 2015 at 1:39 PM, Bair, Raymond A. <[email protected]> wrote:
I do not see any justification for the additional time.
Ray
On 11/19/15, 11:14 AM, "[email protected] on behalf of [email protected]" <[email protected] on behalf of [email protected]> wrote:
Hello,
A change in allocation has been requested:
Requester: redfern (Paul Redfern) Project: Redoxshuttles Title: Redox Shuttles and Additives for Lithium-Ion Batteries Description: Over the past year we have calculated solvation free energies of molecular LiO2 and Li2O2 in various solvents were calculated using various explicit and implicit solvent models, as well as ab initio molecular dynamics (AIMD) methods. Best estimates for the solvation energies from these calculations along with calculated lattice energies of Li2O2 and LiO2 were used to determine the solubility of bulk LiO2 and Li2O2. The solubility of LiO2 was found to be about 17 orders higher than that of Li2O2. The effect of finite crystal size of LiO2 and Li2O2 on the solubility was also considered and was found to increase the solubility, although LiO2 is still much more soluble. The difference in solubilities between LiO2 and Li2O2 will likely affect the growth mechanism and resulting morphologies of the products formed during battery discharge, affecting the performance of the battery cell. We also examined production of a lithium superoxide discharge product on the cathod e in a Li-O2 battery. DFT was used to obtain values of material properties not available experimentally such as LiO2 solubility in DME and elastic properties of the particles. The mesoscale model predicts that coarsening, in which large particles grow and small ones disappear, has a substantial effect on the size distribution of the LiO2 particles during the discharge process. The size evolution during discharge is the result of an interplay between this coarsening process as well as growth. The growth through continued deposition of LiO2 has the effect of causing large particles to grow faster and delays the dissolution of small particles. The predicted size evolution is consistent with experimental results for a previously reported cathode material based on activated carbon during discharge and when it is at rest. The model, even without complex microstructure, can capture size and number of LiO2 particles, but not the shape, e.g. toroid formation. Best estimates for the L iO2 solvation energy from these calculations along with a calculated lattice energy of LiO2 were used to determine the solubility of bulk LiO2. The best estimate for the solubility of LiO2 is about 1 mM, although this has very large uncertainties because small differences in the calculated solvation energy leads to large differences in the solubility due to the exponential. For example, a difference of 2 kcal/mol in solvation energy translates to more than 2 orders of magnitude difference in solubility. The solvation energies are difficult to calculate even to 2 kcal/mol in accuracy. This value from density functional calculations is used initially in the model and subsequently a value for solubility is derived from fits to reproduce experimental data. The solubility from these fits is similar to that predicted from the DFT calculations.
We intend to evaluate at least 200 electrolytes for Lithium-O2 batteries. The most important challenge in Lithium-Air battery research is finding stable non-aqueous electrolytes which permit reversible cycling of the cell. A solvent should have a high dielectric constant in order to dissolve Li+, low viscosity for fast Li+ transport, high stability, low vapor pressure, high oxygen solubility, high O2 and superoxide transport properties, and low toxicity. Solvents should partially dissolve lithium oxide species in order to reduce clogging and increase charge current rate. Quantum chemical calculations can help identify reaction pathways which affect reversibility. So far organic solvents such as ethers, amides, lactams, oxazolidinones, phosphorus compounds and nitriles have been screened for Lithium-Air batteries based on susceptibility to attack by superoxide anion and pKa. No correlation between simple molecular descriptors and activation free energies was found. Low electro philicity has also been used to identify potential candidates, since they would be less susceptible to attack by superoxide or other anion. Other solvents such as sulfones, sulfoxides and ionic liquids have been examined. Lithium salts like LiTFSI can decompose at the lithium electrode and the resulting CF3 radicals can abstract hydrogens from solvents like ethers. Hydrogens on carbons adjacent to heteroatoms (e.g. O, N, F) are more labile due to the anomeric effect. Superoxide anion can abstract protons from proton source impurities (e.g. water) leading to formation of the strong base HOO- via disproportionation which can then react as a nucleophile with the solvent. Superoxide was found not to react with TEGDME ether (Nazar). We will use hydrogen and proton abstraction from and nucleophilic attack by a strong base on the electrolyte along with advanced descriptors (e.g. Fukui functions, hardness and chemical potential) to screen for stable solvents.
Oxygen reduction in the aprotic Li-O2 battery takes place at the boundary between the electrolyte and the cathode so oxygen must dissolve and diffuse in the electrolyte. Unfortunately, aprotic solvents are inadequate in this regard so oxygen enriching materials must be developed. Oxygen enriching materials that have been examined include cobalt porphyrin, iron phthalocyanine, artificial hemoglobin, membranes and perfluorochemicals (PFCs). PFCs have high oxygen solubility, low surface tension and viscosity to enhance O2 diffusivity and Li+ transference number as well as wettability of the cathode surface, high chemical and thermal stability, hydrophobicity and low flammability. In the Li-O2 battery the PFC must be miscible with the polar organic solvent and resistant to nucleophilic attack by superoxide anion radical or HOO-. Miscibility of PFCs with polar organic solvents is increased if LiPFOS is used instead of LiTFSI. Increasing O2 pressure increases O2 diffusivity as well as solubility. We will examine many PFCs for susceptibility to nucleophilic attack by superoxide anion radical and HOO- using DFT calculations. Ab initio calculations will also be done to improve our understanding of interactions between PFCs and O2. Other oxygen enriching materials include reversible oxygen carriers like perfluoro cryptands and crown ethers which are known to have high oxygen carrying capacities, while acyclic perfluoro ethers do not bind O2. DFT calculations will be used to examine perfluoro cryptands and crown ethers. Hybrid ionic liquid fluoro-organic solvent mixtures are also promising as they combine the stability to anion attack of ionic liquids with the high oxygen transport and safety of nonflammable fluoro-organics. We intend to use Gaussian 09 for quantum chemical calculations. In Gaussian 09 the DFT calculations scale well up to 64 processors, however the high-level ab initio methods scale only up to 16 processors. Therefore these calculations do require a longer execution time.
Current: undetermined amount Justification:
Requested: 100000
A specific reason has been given: ran out of time
This needs to be approved and the final allocation amount decided upon.
Thank You, The LCRC Accounts System _______________________________________________ allocations-admins mailing list [email protected] https://lists.lcrc.anl.gov/mailman/listinfo/allocations-admins
_______________________________________________ allocations-admins mailing list [email protected] https://lists.lcrc.anl.gov/mailman/listinfo/allocations-admins
allocations-admins mailing list [email protected] https://lists.lcrc.anl.gov/mailman/listinfo/allocations-admins _______________________________________________ allocations-admins mailing list [email protected] https://lists.lcrc.anl.gov/mailman/listinfo/allocations-admins _______________________________________________ allocations-admins mailing list [email protected] https://lists.lcrc.anl.gov/mailman/listinfo/allocations-admins
_______________________________________________ allocations-admins mailing list [email protected] https://lists.lcrc.anl.gov/mailman/listinfo/allocations-admins
_______________________________________________ allocations-admins mailing list [email protected] https://lists.lcrc.anl.gov/mailman/listinfo/allocations-admins
_______________________________________________ allocations-admins mailing list [email protected] https://lists.lcrc.anl.gov/mailman/listinfo/allocations-admins