All Tutorials

  • Wed May 6, 3:00 pm - 4:30 pm
    575
    Walter Library

    Samuel OMLIN, Ludovic RÄSS, Yuri PODLADCHIKOV

    Scientific Computing Group, Institute of Earth Sciences, University of Lausanne, Switzerland


     

    We present in this talk a pre-release version of HPC.m – the MATLAB HPC-Compiler. We have successively deployed this software in the development of supercomputing applications for Piz Daint in Switzerland (Cray XC30; hybrid with Tesla K20X GPUs) at the Swiss National Supercomputing Centre  (CSCS) in Lugano. We have shown real-world applications work, as for example a non-linear poro-visco-elasto-plastic 3D two-phase solver. The set of non-linear partial differential equations are given below and are solved by time-stepping the equations in time. Our code runs at near peak performance and scales linearly up to more than 2000 GPUs allowing for a spatial resolution of over 20003 grid points. We used explicit time-stepping with sub-iterations in solving the system of time-dependent partial differential equations (see figure 1 and table 1), thus avoiding inversion of large matrices, which would cause severe communication problems. HPC.m transforms simple MATLAB scripts in a few seconds into massively parallel near peak performance applications for CPU-, GPU- and MIC-supercomputers. The MATLAB scripts must employ basic syntax and follow a few simple rules. 

     

    HPC.m enables the following fully automatic steps: (1) translation to a compilable language (to Fortran or C for CPU and MIC, to C-Cuda for GPU), (2) shared memory parallelization (depending on step1: with openMP for Fortran or C, with Cuda for C-Cuda), (3) distributed memory parallelization (with MPI for any of the given options in step 1), (4) deployment of a suitable compiler (depending on the language chosen in step 1). HPC.m is written in MATLAB because the possibility to execute parts of a given input script during code generation allows this task to be better adapted to the given script. HPC.m reduces the time required for transforming a prototype to a near peak performance supercomputing application from typically months or years to a few seconds. 

     

    Key words: HPC.m, Supercomputing, Code-Generation, GPU, MIC, MATLAB

     


  • Tue Apr 21, 1:30 pm - 3:30 pm
    575
    Walter Library

    This tutorial is part of the MSI/NSF Long-read assembly workshop being held on April 9, 2015. Separate registration for the entire event is required. See the description of the event and link to the registration form on the MSI website.


    Recent advances in long read sequencing have revolutionized genome assembly, enabling complete reconstruction of microbial genomes and significantly improved assemblies for higher organisms. This interactive tutorial introduces open source tools and workflows for analyzing, assembling, and interpreting Pacific Biosciences (PacBio) single-molecule real-time (SMRT) sequencing data. The workshop will primarily focus on PacBio’s SMRT Analysis graphical user interface (to be fully supported by MSI), but other command line tools for correction, assembly, and polishing of PacBio data will also be briefly introduced. Topics covered will include managing and importing raw SMRT cell data, adapter removal and subread filtering, SMRT cell loading and quality control, read self-correction and pre-assembly, de novo long read assembly with the hierarchical genome assembly process (HGAP), assembly polishing with Quiver, and base modification and methylation motif detection.


    Basic familiarity with the Unix command line is beneficial but not required. Example data will be provided, but attendees will have the opportunity to analyze their own PacBio data if available.



     


  • Fri Apr 10, 10:00 am - 12:00 pm
    138
    Cargill Building

    This tutorial is part of the MSI/NSF Long-read assembly workshop being held on April 9, 2015. Separate registration for the entire event is required. See the description of the event and link to the registration form on the MSI website.


    Recent advances in long read sequencing have revolutionized genome assembly, enabling complete reconstruction of microbial genomes and significantly improved assemblies for higher organisms. This interactive tutorial introduces open source tools and workflows for analyzing, assembling, and interpreting Pacific Biosciences (PacBio) single-molecule real-time (SMRT) sequencing data. The workshop will primarily focus on PacBio’s SMRT Analysis graphical user interface (to be fully supported by MSI), but other command line tools for correction, assembly, and polishing of PacBio data will also be briefly introduced. Topics covered will include managing and importing raw SMRT cell data, adapter removal and subread filtering, SMRT cell loading and quality control, read self-correction and pre-assembly, de novo long read assembly with the hierarchical genome assembly process (HGAP), assembly polishing with Quiver, and base modification and methylation motif detection.


    Basic familiarity with the Unix command line is beneficial but not required. Example data will be provided, but attendees will have the opportunity to analyze their own PacBio data if available.



     


  • Thu Apr 9, 9:00 am - 11:30 am
    138
    Cargill Building

    This tutorial is part of the MSI/NSF Long-read assembly workshop being held on April 9, 2015. Separate registration for the entire event is required. See the description of the event and link to the registration form on the MSI website.


    Recent advances in long read sequencing have revolutionized genome assembly, enabling complete reconstruction of microbial genomes and significantly improved assemblies for higher organisms. This interactive tutorial introduces open source tools and workflows for analyzing, assembling, and interpreting Pacific Biosciences (PacBio) single-molecule real-time (SMRT) sequencing data. The workshop will primarily focus on PacBio’s SMRT Analysis graphical user interface (to be fully supported by MSI), but other command line tools for correction, assembly, and polishing of PacBio data will also be briefly introduced.

    Topics covered will include managing and importing raw SMRT cell data, adapter removal and subread filtering, SMRT cell loading and quality control, read self-correction and pre-assembly, de novo long read assembly with the hierarchical genome assembly process (HGAP), assembly polishing with Quiver, and base modification and methylation motif detection.


    Basic familiarity with the Unix command line is beneficial but not required. Example data will be provided, but attendees will have the opportunity to analyze their own PacBio data if available.


  • Fri Apr 3, 1:00 pm - 3:00 pm
    575
    Walter Library

    Progenesis QI staff will give a hands-on tutorial demonstrating Progenesis QI's analysis capabilities. 


  • Fri Apr 3, 9:00 am - 12:00 pm

    Staff from Progenesis QI will give an overview of what Progenesis QI can do, what is new in version 2 and to answer your questions. 


    Morning Lecture will be in Mayo 3-125
    Afternoon Hands-on will be in Walter Library 575


    Schedule:
    9:00 am registration 
    9:30 am introduction
    9:45 am Progenesis QI overview and what’s new in Version 2
    10:00 – 11:50 am Progenesis QI operational overview with questions
    12:00 pm Lunch on the 5th floor of Walter Library
    1:00 – 3:30 pm hands on demonstration in Walter Library 575


  • Thu Mar 19, 1:30 pm - 4:30 pm
    138
    Cargill Building

    -


  • Tue Jan 13, 10:00 am - 11:30 am
    138
    Cargill Building

    This tutorial is part of the MPGI/MSI PacBio workshop being held on December 10, 2014. Separate registration for the entire event is required. See the description of the event and link to the registration form on the MSI website.


    Recent advances in long read sequencing have revolutionized genome assembly, enabling complete reconstruction of microbial genomes and significantly improved assemblies for higher organisms. This interactive tutorial introduces open source tools and workflows for analyzing, assembling, and interpreting Pacific Biosciences (PacBio) single-molecule real-time (SMRT) sequencing data. The workshop will primarily focus on PacBio’s SMRT Analysis graphical user interface (to be fully supported by MSI), but other command line tools for correction, assembly, and polishing of PacBio data will also be briefly introduced.


    Topics covered will include managing and importing raw SMRT cell data, adapter removal and subread filtering, SMRT cell loading and quality control, read self-correction and pre-assembly, de novo long read assembly with the hierarchical genome assembly process (HGAP), assembly polishing with Quiver, and base modification and methylation motif detection.


    Basic familiarity with the Unix command line is beneficial but not required. Example data will be provided, but attendees will have the opportunity to analyze their own PacBio data if available.


  • Wed Jan 7, 1:00 pm - 2:30 pm
    138
    Cargill Building

    This tutorial is part of the MPGI/MSI PacBio workshop being held on December 10, 2014. Separate registration for the entire event is required. See the description of the event and link to the registration form on the MSI website.


    Recent advances in long read sequencing have revolutionized genome assembly, enabling complete reconstruction of microbial genomes and significantly improved assemblies for higher organisms. This interactive tutorial introduces open source tools and workflows for analyzing, assembling, and interpreting Pacific Biosciences (PacBio) single-molecule real-time (SMRT) sequencing data. The workshop will primarily focus on PacBio’s SMRT Analysis graphical user interface (to be fully supported by MSI), but other command line tools for correction, assembly, and polishing of PacBio data will also be briefly introduced.


    Topics covered will include managing and importing raw SMRT cell data, adapter removal and subread filtering, SMRT cell loading and quality control, read self-correction and pre-assembly, de novo long read assembly with the hierarchical genome assembly process (HGAP), assembly polishing with Quiver, and base modification and methylation motif detection.


    Basic familiarity with the Unix command line is beneficial but not required. Example data will be provided, but attendees will have the opportunity to analyze their own PacBio data if available.


  • Wed Dec 17, 1:00 pm - 2:30 pm
    138
    Cargill Building

    This tutorial is part of the MPGI/MSI PacBio workshop being held on December 10, 2014. Separate registration for the entire event is required. See the description of the event and link to the registration form on the MSI website.


    Recent advances in long read sequencing have revolutionized genome assembly, enabling complete reconstruction of microbial genomes and significantly improved assemblies for higher organisms. This interactive tutorial introduces open source tools and workflows for analyzing, assembling, and interpreting Pacific Biosciences (PacBio) single-molecule real-time (SMRT) sequencing data. The workshop will primarily focus on PacBio’s SMRT Analysis graphical user interface (to be fully supported by MSI), but other command line tools for correction, assembly, and polishing of PacBio data will also be briefly introduced.


    Topics covered will include managing and importing raw SMRT cell data, adapter removal and subread filtering, SMRT cell loading and quality control, read self-correction and pre-assembly, de novo long read assembly with the hierarchical genome assembly process (HGAP), assembly polishing with Quiver, and base modification and methylation motif detection.


    Basic familiarity with the Unix command line is beneficial but not required. Example data will be provided, but attendees will have the opportunity to analyze their own PacBio data if available.


Pages