SciNet User Support Library

Welcome to the SciNet User Support library. Here you will find up-to-date manuals put together by SciNet staff and users, as well as links to external resources, to help you make use of SciNet resources for computational scientific discovery.

SciNet System Status System Status System status can now be found at docs.scinet.utoronto.ca Mon 23 Apr 2018 GPC-compute is decommissioned, GPC-storage available until 30 May 2018 Thu 18 Apr 2018 Niagara system will undergo an upgrade to its Infiniband network between 9am and 12pm, should be transparent to users, however there is a chance of network interruption. Fri 13 Apr 2018 HPSS system will be down for a few hours on Mon, Apr/16, 9AM, for hardware upgrades, in preparation for the eventual move to the Niagara side. Tue 10 Apr 2018 Niagara is open to users. Wed 4 Apr 2018 We are very close to the production launch of Niagara, the new system installed at SciNet. While the RAC allocation year officially starts today, April 4/18, the Niagara system is still undergoing some final tuning and software updates, so the plan is to officially open it to users on next week. All active GPC users will have their accounts, $HOME, and $PROJECT, transferred to the new Niagara system. Those of you who are new to SciNet, but got RAC allocations on Niagara, will have your accounts created and ready for you to login. We are planning an extended Intro to SciNet/Niagara session, available in person at our office, and webcast on Vidyo and possibly other means, on Wednesday April 11 at noon EST.
QuickStart Guides Essential Login Information GPC - General Purpose Cluster: Nehalem (x86_64) cluster running Linux TCS - Tightly Coupled System: IBM Power 6 system running AIX Storage at SciNet Data Transfer from SciNet Software & Libraries Job Scheduling System (Moab) A Performance Primer Frequently Asked Questions Performance & Debugging Tools: GPC Performance & Debugging Tools: TCS Acknowledging SciNet SciNet Accounts Application Process SciNet Usage Policy	Manuals Intel Manuals (GPC): C/C++: Intel (R) C++ Compiler 11.1 User and Reference Guide FORTRAN: Intel(R) Fortran Compiler 11.1 User and Reference Guide MKL: Intel(R) Math Kernel Library for Linux* OS User's Guide, Intel(R) Math Kernel Library Link Line Advisor MPI: Intel(R) MPI Library for Linux* OS Reference Manual IBM AIX Manuals (TCS): C: Compiler Reference, Optimization and Programming Guide, Language Reference. C++: Compiler Reference, Optimization and Programming Guide, Language Reference. FORTRAN: Compiler Reference, Optimization and Programming Guide, Language Reference. ESSL (High performance math library): Guide and Reference, Parallel ESSL Guide and Reference Application Performance Tuning and Optimization on POWER6 LoadLeveler: LoadLeveler Using and Administering Parallel Environment: Parallel Environment Operation and Use Adaptive Computing (Cluster Resources) Manuals: MOAB: Moab Workload Manager TORQUE: Torque Resource Manager
Tutorials SciNet User Tutorial Workshops given at SciNet: SciNet Parallel Scientific Computing Workshop, Aug 2009 - Slides and Video IBM AIX Workshop, Nov 2008 IBM MPI Workshop, Feb 2009 IBM Compiler Workshop, Feb 2009 Online Parallel/Scientific Programming Courses Virtual School for CSE -- Free online webinar-style courses, but one must register beforehand: Petascale Programming Environments and Tools, July 6–9, 2010 Big Data for Science, July 26–30, 2010 Proven Algorithmic Techniques for Many-core Processors, August 2–6, 2010 NVidia archived courses for GPGPU Programming Useful Links: LLNL MPI Tutorial - This tutorial by Blaise Barney was used as the basis for Kirk Jordan's MPI workshop at SciNet. It is well worth studying. High Performance Computing Book - An online version of an older O'Reilly book which covers the basics of (mostly serial) programming for performance. The most important issues today are covered here, and very clearly. How to Write Fast Numerical Code - a good introduction to thinking about performance.
User-Supplied Content Share your expertise with the SciNet Community! Running specific codes on SciNet machines Tips for performance or convenience on SciNet machines Running serial batch jobs (including GNU parallel) Using the ramdisk on SciNet's GPC User-space perl modules Packing up large data sets and putting them on HPSS Remote batch