S Jarp

Publications (14)0 Total impact

• Conference Proceeding: Evaluation of Likelihood Functions for Data Analysis on Graphics Processing Units

  S. Jarp, A. Lazzaro, J. Leduc, A. Nowak, F. Pantaleo
  [show abstract] [hide abstract]
  ABSTRACT: Data analysis techniques based on likelihood function calculation play a crucial role in many High Energy Physics measurements. Depending on the complexity of the models used in the analyses, with several free parameters, many independent variables, large data samples, and complex functions, the calculation of the likelihood functions can require a long CPU execution time. In the past, the continuous gain in performance for each single CPU core kept pace with the increase on the complexity of the analyses, maintaining reasonable the execution time of the sequential software applications. Nowadays, the performance for single cores is not increasing as in the past, while the complexity of the analyses has grown significantly in the Large Hadron Collider era. In this context a breakthrough is represented by the increase of the number of computational cores per computational node. This allows to speed up the execution of the applications, redesigning them with parallelization paradigms. The likelihood function evaluation can be parallelized using data and task parallelism, which are suitable for CPUs and GPUs (Graphics Processing Units), respectively. In this paper we show how the likelihood function evaluation has been parallelized on GPUs. We describe the implemented algorithm and we give some performance results when running typical models used in High Energy Physics measurements. In our implementation we achieve a good scaling with respect to the number of events of the data samples.
  Parallel and Distributed Processing Workshops and Phd Forum (IPDPSW), 2011 IEEE International Symposium on; 06/2011
• Article: Perfmon2: a leap forward in performance monitoring

  S Jarp, R Jurga, A Nowak
  [show abstract] [hide abstract]
  ABSTRACT: This paper describes the software component, perfmon2, that is about to be added to the Linux kernel as the standard interface to the Performance Monitoring Unit (PMU) on common processors, including x86 (AMD and Intel), Sun SPARC, MIPS, IBM Power and Intel Itanium. It also describes a set of tools for doing performance monitoring in practice and details how the CERN openlab team has participated in the testing and development of these tools.
  Journal of Physics Conference Series 07/2008; 119(4):042017.
• Article: Helping to choose the right commodity compilers for high energy physics

  N Buncic, S Jarp, C. Meessen, H Tang, R Yaari
• Article: The breaking point of modern processor and platform technology

  A Nowak, S Jarp, A Lazzaro, J Leduc
  [show abstract] [hide abstract]
  ABSTRACT: This work is an overview of state of the art processors used in High Energy Physics, their architecture and an extensive outline of the forthcoming technologies. Silicon process science and hardware design are making constant and rapid progress, and a solid grasp of these developments is imperative to the understanding of their possible future applications, which might include software strategy, optimizations, computing center operations and hardware acquisitions. In particular, the current issue of software and platform scalability is becoming more and more noticeable, and will develop in the near future with the growing core count of single chips and the approach of certain x86 architectural limits. Other topics brought forward include the hard, physical limits of innovation, the applicability of tried and tested computing formulas to modern technologies, as well as an analysis of viable alternate choices for continued development.
• Article: First I.N.F.N. International School on "Architectures, tools and methodologies for developing efficient large scale scientific computing applications"

  F Ruggieri, P. Calafiura, A. De Salvo, P Elmer, F Furano, S Jarp, F Galeazzi, A Hanushevsky, V. Innocente, A Lazzaro, U. Marconi, P Mato, M Morandin, A Nowak, M. Paterno, D Salomoni, R Stroili, L. Tuura, V. Vagnoni
• Article: Second I.N.F.N. International School on "Architectures, tools and methodologies for developing efficient large scale scientific computing applications"

  F Ruggieri, S Binet, P. Calafiura, P Elmer, F Furano, D. Galli, S Jarp, A Hanushevsky, V. Innocente, A Lazzaro, P Mato, A Nowak, D Salomoni, L. Tuura, V. Vagnoni
• Source

  Available from: cern.ch

  Article: PC as physics computer for LHC?

  S Jarp, H Tang, A Simmins, R Yaari
  [show abstract] [hide abstract]
  ABSTRACT: In the last five years, we have seen RISC workstations take over the computing scene that was once controlled by mainframes and supercomputers. In this paper we will argue that the same phenomenon might happen again. A project, active since March this year in the Physics Data Processing group of CERN's CN division is described where ordinary desktop PCs running Windows (NT and 3.11) have been used for creating an environment for running large LHC batch jobs (initially the DICE simulation job of Atlas). The problems encountered in porting both the CERN library and the specific Atlas codes are described together with some encouraging benchmark results when comparing to existing RISC workstations in use by the Atlas collaboration. The issues of establishing the batch environment (Batch monitor, staging software, etc.) are also covered. Finally a quick extrapolation of commodity computing power available in the future is touched upon to indicate what kind of cost envelope could be sufficient for the simulation farms required by the LHC experiments.
• Source

  Available from: cern.ch

  Article: Parallelization of maximum likelihood fits with OpenMP and CUDA

  S Jarp, A Lazzaro, J Leduc, A Nowak, F Pantaleo
  [show abstract] [hide abstract]
  ABSTRACT: Data analyses based on maximum likelihood fits are commonly used in the high energy physics community for fitting statistical models to data samples. This technique requires the numerical minimization of the negative log-likelihood function. MINUIT is the most common package used for this purpose in the high energy physics community. The main algorithm in this package, MIGRAD, searches the minimum by using the gradient information. The procedure requires several evaluations of the function, depending on the number of free parameters and their initial values. The whole procedure can be very CPU-time consuming in case of complex functions, with several free parameters, many independent variables and large data samples. Therefore, it becomes particularly important to speed-up the evaluation of the negative log-likelihood function. In this paper we present an algorithm and its implementation which benefits from data vectorization and parallelization (based on OpenMP) and which was also ported to Graphics Processing Units using CUDA.
• Source

  Available from: cern.ch

  Article: Performance awareness: execution performance of HEP codes on RISC platforms,issues and solutions

  R Yaari, S Jarp
  [show abstract] [hide abstract]
  ABSTRACT: The work described in this paper was started during the migration of Aleph's production jobs from the IBM mainframe/CRAY supercomputer to several RISC/Unix workstation platforms. The aim was to understand why Aleph did not obtain the performance on the RISC platforms that was "promised" after a CERN Unit comparison between these RISC platforms and the IBM mainframe. Remedies were also sought. Since the work with the Aleph jobs in turn led to the related task of understanding compilers and their options, the conditions under which the CERN benchmarks (and other benchmarks) were run, kernel routines and frequently used CERNLIB routines, the whole undertaking expanded to try to look at all the factors that influence the performance of High Energy Physics (HEP) jobs in general. Finally, key performance issues were reviewed against the programs of one of the LHC collaborations (Atlas) with the hope that the conclusions would be of long- term interest during the establishment of their simulation, reconstruction and analysis codes.
• Source

  Available from: cern.ch

  Article: The ATLAS Computing Model

  D Adams, D Barberis, C P Bee, R Hawkings, S Jarp, R Jones, D. Malon, L. Poggioli, G Poulard, D Quarrie, T. Wenaus
  [show abstract] [hide abstract]
  ABSTRACT: The ATLAS Offline Computing Model is described. The main emphasis is on the steady state, when normal running is established. The data flow from the output of the ATLAS trigger system through processing and analysis stages is analysed, in order to estimate the computing resources, in terms of CPU power, disk and tape storage and network bandwidth, which will be necessary to guarantee speedy access to ATLAS data to all members of the Collaboration. Data Challenges and the commissioning runs are used to prototype the Computing Model and test the infrastructure before the start of LHC operation. The initial planning for the early stages of data-taking is also presented. In this phase, a greater degree of access to the unprocessed or partially processed raw data is envisaged.
• Article: Evaluating the scalability of HEP software and multi-core hardware

  S Jarp, A Lazzaro, J Leduc, A Nowak
  [show abstract] [hide abstract]
  ABSTRACT: As researchers have reached the practical limits of processor performance improvements by frequency scaling, it is clear that the future of computing lies in the effective utilization of parallel and multi-core architectures. Since this significant change in computing is well underway, it is vital for HEP programmers to understand the scalability of their software on modern hardware and the opportunities for potential improvements. This work aims to quantify the benefit of new mainstream architectures to the HEP community through practical benchmarking on recent hardware solutions, including the usage of parallelized HEP applications.
• Article: The ATLAS Experiment at the CERN Large Hadron Collider

  G. Aad, E. Abat, J Abdallah, AA Abdelalim, A. Abdesselam, O. Abdinov, BA Abi, M Abolins, H Abramowicz, E. Acerbi, [......], G. Unel, Y. Unno, E. Urkovsky, G. Usai, Y. Usov, L Vacavant, V Vacek, B Vachon, S. Vahsen, C. Valderanis
• Article: ATLAS computing technical proposal

  ATLAS Collaboration, A. Airapetian, V. Grabsky, H. Hakopian, A Vartapetian, F Fares, GF Moorhead, ME Sevior, GN Taylor, SN Tovey, [......], T Zhao, D Fasching, S Gonzalez, RC Jared, YB Pan, IJ Scott, J Walsh, SL Wu, JM Yamartino, G Zobernig
• Article: ATLAS calorimeter performance

  ATLAS Collaboration, A. Airapetian, V. Grabsky, H. Hakopian, A Vartapetian, F Fares, GF Moorhead, ME Sevior, GN Taylor, SN Tovey, [......], T Zhao, D Fasching, S Gonzalez, R C Jared, YB Pan, IJ Scott, J Walsh, SL Wu, JM Yamartino, G Zobernig