# What / Who is ResearchIT at Iowa State? * What do we do? * Collaborate with researchers to solve complex technology problems, freeing them up to focus on their research * Provide access to computational hardware, software, storage, consulting * Focus our time on tasks that scale * Who? * Collaboration with LAS, CALS, ITS * [http://researchit.las.iastate.edu/people](http://rit.las.iastate.edu/people) * Lots of involvement from other IT people in college, departments, and ITS * researchit@iastate.edu --- # Pricing Model * We do not charge for our time * Our salaries are paid through the colleges, departments, etc. that we support * Offsetting liability * We do not charge per CPU cycle * We want to collaborate with your PI's on grant proposals to include funding for access to shared infrastructure * We will do our best to find spare cycles to run your research if you don't have funding (may not be on the newest hardware) --- # Data Storage * Large Scale Storage (aka Research Files) * Stable, inexpensive storage for large research data * Located in Durham Data Center * File snapshots: hourly, daily, weekly, monthly * Fast network access (40Gbps x 2) * LAS & CALS provide each PI 2TB of free storage * PI cost for additional storage is $40 / TB / Year (cheaper than Amazon Glacier) * For more info on LSS: [https://researchit.las.iastate.edu/large-scale-storage-lss](https://researchit.las.iastate.edu/large-scale-storage-lss) * Please do not store your critical research data on your laptop or on external USB disk drives --- # [Computational Resources](https://rit.las.iastate.edu/hardware) - Custom Machines * RIT1 / RIT2: * Older machines, free for research use. All bcbgrads are on the list for rit1. * BigRam: * for when your whole dataset needs to fit in RAM (e.g. de novo assembly) * BioCrunch, Biocrunch2, Biocrunch3: * for multithreaded, non MPI software (mapping, alignment, etc.) * Speedy / Speedy2: * Fewer faster cores * For single threaded software * Legion: * 2176 threads across 8 nodes * Meant for highly parallel workloads (molecular dynamics, genome alignment, Monte Carlo simulations) * GPU01: * 8 NVIDIA 1080Ti * Molecular Dynamics, CyroEM, Deep Learning, etc. * LAS-Win-01: * When software will only run on Windows --- # Computational Resources - University Clusters * [Condo 2017 HPC Cluster](http://www.hpc.iastate.edu/guides/condo): * Approx 200 Nodes * Primarly for MPI capable software * For sponsored research (College / PI supported) * [Condo 2017 Free Tier](http://www.hpc.iastate.edu/guides/condo): * 48 nodes + a few specialty nodes * This tier is to allow groups without funding to complete research, gain an understanding of the HPC environment, etc. * [hpc-class](http://www.hpc.iastate.edu/guides/classroom-hpc-cluster) * For classes, not research * New cluster coming later this fall - details coming soon --- # Other Research Computing Resources ## [Xsede](https://www.xsede.org/) * ISU researchers have access to the national supercompute centers (e.g. TACC, PSC) via Xsede * For problems that need to scale larger than our on-campus resources * Contact campus champoins: Levi Baber, Jim Coyle, or Andrew Severin ## Cloud (AWS, Azure, etc.) * Tempting introductory rates * Be cautious of putting in a credit card (charges can accumulate quickly) * Consider data transfer times and speed * Consult with IT before purchasing - we have special negotiated rates --- # Older Compute Resources at ISU * cyence * Being merged / repurposed * cystorm * Being retired * lightning3 * Being retired --- # A note on operating systems: * High Performance Computing systems you encounter will typically be running Linux * At ISU that will be Red Hat Enterprise Linux (RHEL) * [Fedora](https://getfedora.org/) is the upstream desktop distribution (good option to become familiar with linux) * [CentOS](https://www.centos.org/) is the downstream server distribution * When you google linux questions, you'll inevitably get results for Ubuntu * results can point you in the right direction, but interpret and understand - don't just copy and paste * sudo apt-get install - will not work for you on our systems --- # Software ## Modules * We use modules to allow multiple people to use the same software with consistent, reproducible results * Modules keep your environment clean and free of conflicts (e.g. python2/3 or Java7/8) * Think of modules as a software library. You need to check out the software before you can use it. * We use [Spack](https://researchit.las.iastate.edu/research-it-software-archive) to build our software, and sync it to hpc-class, condo, and all of our custom machines ``` $ module avail ----------------------------------- /opt/rit/spack-modules/lmod/linux-rhel7-x86_64/Core ----------------------------------- abyss/2.0.2-py2-cuda9-openmpi3-o6zuizq netlib-scalapack/2.0.2-openmpi3-sakcdle (D) r-evaluate/0.10.1-py2-r3.4-7wldk4v aegean/0.15.2-py2-dwjwigu nettle/3.3-jcgc7wh r-fastmatch/1.1-0-py2-r3.4-l7jos6a albert/4.0a_opt4-dtcanam nnvm/20170418-ojgzdup r-flexclust/1.3-5-py2-r3.4-tjnam64 ... ``` ## BioDatabases on Condo at : /work/LAS/BioDatabase * BLASTdb, FASTA, iGenomes, mpiBLASTdb --- # Software ## Modules * To use a module ``` $ module load abyss ``` * Default behavior is to load the latest version if not specified * Use 'module purge' to clear your environment before loading something different --- # Consulting * Our goal is to work on tasks that can benefit others through packaged software, documentation, collaboration, etc. and that help you overcome technical challenges that moves your science forward and enables faster iteration * Can advise on computational performance of software * Can help you identify optimal software for your pipeline * Help identify ways to optimize pipeline * Hardware * Different software * Parallelization wrappers (e.g. gnu parallel) --- # Consulting Examples * Hufford lab cut genome assembly time by running on BigRAM * Helped Nikolau lab migrate from MEGA to RAxML * Sivasankar lab cut MD (gromacs) job run time, and costs by moving off Amazon EC2 to Condo * Helped Windus lab compile GAMESS on Condo * Helped Dobbs lab publish web interface for rpimotif tool * Helped Wolter lab parallelize matlab code * Helped Lu lab port code from Windows to Linux --- # Websites * Lab Websites * Custom web based research tools * Secure virtual web environments * Code review * Our platform: http://www.biology-it.iastate.edu/luggage_doc/ * Need to publish a new service, or tool on the web ? * [http://it.las.iastate.edu/public-ip-or-dns-request](http://it.las.iastate.edu/public-ip-or-dns-request) --- # Other Services * Amazon Web Services: * e.g. S3 buckets to host bigwig files for gbrowse, EC2 instances for compute jobs, GPU nodes * Version Control: * Can help you start tracking code in git, several options for hosting including on-site: http://it.las.iastate.edu/version-control * Backups: * We can set you up with backup software to protect data on your local disk: http://it.las.iastate.edu/computer-backup --- # Tutorials * Videos: [http://rit.las.iastate.edu/video](https://rit.las.iastate.edu/video) * How To: [https://researchit.las.iastate.edu/category/how-to](https://researchit.las.iastate.edu/category/how-to) --- # Support * [Research IT](http://rit.las.iastate.edu) * researchit@iastate.edu * IRC (chat): #bitcom on freenode * [ISU IT HPC Team](http://hpc.iastate.edu) * Condo & HPC-Class Support team * hpc-help@iastate.edu