Simulation Project Leaders Meeting 8-May-2007 ==== Attendees John Apostolakis (JA), Witek Pokorski (WP), Alberto Ribon (AR), Gabriele Cosmo (GC). Excused: Alfredo Ferrari (AF), Pere Mato (PM), Rob Veenhof (RV). --------------------------------------------------------------------- ==== Minutes - CHEP07: o An abstract has been sent by M.Gallas for Flugg and recent application in physics validation but accepted only as poster, it may be withdrawn in case not accepted as presentation. o The abstract on the status of Geant4 electromagnetic physics by V.Ivantchenko has been accepted. o The abstract about the status of the GENSER project by M.Kirsanov has been accepted. - Technical students: o One technical student has been assigned for work in GDML/FLUGG. --------------------------------------------------------------------- ==== Simulation physics validation - Alberto - Meetings: o The last Physics Validation meeting was on March 28th. There has been a presentation on the development in quasi-elastic processes by M.Kossov, development which will be included in 8.3 with expected impact on the simulation of the longitudinal shower profiles. Another presentation from ATLAS by T.Carli on the combined test-beam analysis, where the effects of Birks quenching has been discussed and no big effect was observed. o At the next meeting, there will be two presentations from CMS on test-beam validation of electromagnetic and hadronic physics of Geant4. - Simple benchmarks: o It is not yet decided which benchmark to prioritize for the next study. Thin-target benchmark on single diffraction from Helios may be the candidate. o An additional benchmark focused to Geant4 could be also considered. o A working plan for the next 12 months will be soon provided, which will include also the benchmarks to be considered. - Extension of ATLAS TileCal 2002 analysis with Fluka: o Preliminary results of the analysis are ready. There has been a meeting with Geant4 experts last week where the results have been discussed; on Thursday May 10th a meeting will take place also with the Fluka experts. - Shower shapes: o The note on the shower shapes study has been published and is available from the web: http://lcgapp.cern.ch/project/mgmt/doc.html. - Correction of test-beam data: o No activity at the moment. Most people busy in finalising the test beam analyses in the experiments. --------------------------------------------------------------------- ==== Generator services - Witek - Meetings: o The Generators Services monthly meeting took place on April 4th; the status of GENSER was presented by M.Kirsanov; the Generators FAQ page was demonstrated by B.Lloyd; S.Gieseke (currently at CERN as part of SFT group) gave a talk on the status of Herwig++. o The next Generators Services monthly meeting is scheduled for June 6th. o On May 25th the second Generator Services project planning meeting will take place. WP will present what has been done during last six months. The future goals of the project will be also discussed. - GENSER: o A.Polaryush has finished his stay at CERN; after a period of training, he contributed with one new test for Pythia. At present, A.Toropin is at CERN; also for him this isis his first visit since all the changes introduced in GENSER, therefore some training is required as well. Due to this, progress in the implementation of the new tests has been slow in the last month. The list of generators is, however, complete and experiments are using it. o Generators FAQ: Generator authors have been invited to use it and provide feedback. No much activity so far, contributions received from M.Mangano and T.Sjostrand. - There has been a request from ATLAS to take over the maintenance of the ATLAS interface package for Photos (in response to an enquiry by WP concerning more direct contribution of the project to the experiments). WP is currently getting familiar with the package and the issues that need to be solved there. In the same light, it was also discussed again with CMS about possible contribution; CMS will propose some concrete test/validation to be performed. - L.Sonnenschein has started working on using Root-I/O for HepMC2; one technical problem was identified in Reflex; currently looking for a solution. --------------------------------------------------------------------- ==== Fluka - Alfredo - No report. --------------------------------------------------------------------- ==== Simulation framework - Witek - Flugg: o The package has been moved to the new Fluka; there have been problems which have been reported to AF; a few days later a patch version of Fluka was released. Flugg is now working fine with the latest Fluka. o GDML: o B.Lloyd has finished his technical studenship. A number of things still need to be fixed (especially in GDML->Root converter), users are submitting bug reports and there's the need for someone to take it over quickly. The integration of the GDML plugin in Geant4 was not completed by B.Llyod. --------------------------------------------------------------------- ==== Geant4 - John - Geant4 review: o The Geant4 public review was held at CERN on April 16-20th. The focus was on physics performance, CPU performance and usability; there were presentations of results and issues from HEP experiments and other users and reports and responses from developers. o Concerning CPU performance, users expressed strong interest in obtaining improved CPU performance. Providing support for users to enable them to realise performance improvement with existing releases and realising kernel CPU improvements, were both considered essential elements for achieving this. Creating new variance reduction methods, in kernel or user code, was seen as an alternative avenue for appropriate applications. o Significant interest of users existed for simple documentation of the physics choices available in existing physics lists. Clear choices to enable the user choose between accuracy of physics response and CPU performance were requested. o Several users also expressed interest in having additional physics models. o The reviewers noted positively the large effort made for validation and verification during the past year. A summary of comparison results for thin targets and thicker setups was presented by developers in Geant4. o A clear request was made by users polled to provide more information in the release notes concerning details/effects of fixes and developments and provide clear information on expected physics and CPU performance. - Benchmarks: o A new set of benchmarks has been put in place. 5 setups: pure geometry, geometry and field, EM-physics, hadronic physics and full detector setups using GDML. A CMS GDML file was created by A.Novak, W.Pokorski and A.Ribon. - Releases: o Release 8.3 will be announced this week. It includes a revised Fritiof FTF model, and in QGS physics lists a separation of the inelastic cross-section between a new quasi-elastic hadronic model and the existing QGS model. A number of other fixes and improvements are also included. Improvements in the energy-loss process provide better stability when varying production thresholds. A candidate tag of the release was made available to experiments on April 27th; results of benchmarking on ATLAS have been received, reporting an issue with CPU performance (20% penalty compared with 7.1.p01 using QGSP_EMV); the issue is under investigation on both sides. Other issues addressed in release 8.3 are fixes for NaNs in various physics models, and CPU benchmarks. - Shower shapes: o Release 8.3 has been tested in simplified calorimeter setups; the shower shapes were longer for the revised FTF physics lists and the QGS physics lists tested including QGSP_EMV, QGSP. --------------------------------------------------------------------- ==== Garfield - Rob - No report. --------------------------------------------------------------------- ==== AOB Next meeting scheduled for June 19th.