Computing Facilities

Most computational groups at the University of Toronto are provided with High Performance Computing (HPC) facilities in coordination with SciNet and Compute/Calcul Canada.

Large scale computations are made possible through High Performance Computing (HPC) facilities. The University of Toronto's primary computing facility is Scinet, Canada's largest supercomputer centre, which also provides resources for Compute/Calcul Canada, a national infrastructure of for supercomputing research. SciNet was established in 2009 as a consortium of the University of Toronto and its affiliated research hospitals. A large spectrum of computational research takes place on SciNet including astrophysics, astronomy, fluid dynamics, structures, biomedical engineering, material science, electrical engineering and many more. In addition to providing computing resources and technical support, SciNet also offers a wide variety of courses and specialized training. For more information, visit Scinet's training and education page.

Systems

Multiple systems have been installed at the SciNet Centre. A brief description of a few select sytems is given below.

General Purpose Cluster (GPC)

The General Purpose Cluster is a large cluster consisting of 3,780 nodes with 16 GB of memory or more and 8 Intel cores each. The nodes use a shared parallel file system and have no local disks. The compute nodes are accessed through a queuing system that allows jobs with a minimum of 15 minutes and maximum wall time of 48 hours . The GPC is an IBM iDataPlex cluster based on Intel's Nehalem architecture (one of the first in the world to make use of the new chips). The GPC consists of 3,780 nodes (IBM iDataPlex DX360M2) with a total of 30,912 cores (Intel Xeon E5540) at 2.53GHz, with 16GB RAM per node (2GB per core). Approximately one quarter of the cluster is interconnected with non-blocking DDR InfiniBand while the rest of the nodes are connected with 5:1 blocked QDR InfiniBand. From an operational standpoint, the contributed systems Gravity and Sandy, as well as the Viz system, are more or less part of the GPC. For more information, visit the GPC quickstart wiki page.

Tightly Coupled System (TCS)

The TCS is a cluster of high memory, many-core IBM Power 575 nodes, each with 4.7GHz Power 6 processors on a very fast infiniband interconnect. At ~3,000 cores it is relatively small when compared to the GPC, and is dedicated for jobs requiring a large memory/low latency configuration. Each node contains 16 processors, each with 2 independant cores. At least 32 tasks are required to fully utlize a single node. For more information, visit the TCS quickstart wiki page.

Sandybridge (Sandy)

The Sandy cluster consists of 76 x86_64 nodes with two octal Intel Xeon E5-2650 2.0GHz CPUs with 64GB of RAM per node. The nodes are interconnected with 2.6:1 blocking QDR Infiniband for MPI communications and disk I/O to the SciNet GPFS filesystems. In total this cluster contains 1216 x86_64 cores with 4,864 GB of total RAM. Sandy is a user-contributed system. For more information, visit the Sandy quickstart wiki page.