Imagine if you gave every single student, professor, lecturer, administrator and support staff member at Carleton University a new MacBook.
 
Then imagine what would happen if you asked them to hook all the computers up together 
and start furiously doing calculations.
 
That’s nearly 30,000 people working in concert with computers faster than most of us need on a day-to-day basis.
 
But that’s still not as powerful as the University of Toronto’s new supercomputer system.
 
This system is operated by SciNet and was recently declared the most powerful supercomputer in Canada and the 16th most powerful in the world.
 
It’s used for research in anything from biology to high-energy physics to climate change studies to language analysis, said Richard Peltier, scientific director of SciNet and U of T physics professor.
 
The SciNet supercomputer, called the General Purpose Cluster, is composed of 30,240 nodes connected together.
Each node is like an individual computer, with a 2.5 gigahertz processor and 2 gigabytes of RAM.
 
In addition to the large model, SciNet runs the Tightly Coupled System (TCS), a smaller supercomputer –— the second most powerful in Canada — that is the sprinter to the large model’s endurance runner.
 
We think of [the TCS] sort of as a race car. It’s not very efficient to run, it’s not lean on gas, but it goes very, very fast,” Peltier said.
 
The computer is housed in 
a large warehouse, which is divided into a room-within-a-room, Peltier said.
 
The inner room is where the supercomputer is set up, and the outer room contains the massive cooling system needed to keep 
the supercomputer running and not overheat.
 
The warehouse is located in the town of Vaughan, Ont., just north of Toronto.
 
Originally, SciNet wanted to build the supercomputer in downtown Toronto near U of T’s campus, Peltier said.
 
But because of the supercomputer’s very large power requirements, there wasn’t enough spare electricity in downtown Toronto to run it.
 
If they had put it downtown, a new power substation would have had to be built in the city, Peltier said.
 
Most of SciNet’s computing power will be given out to other researchers outside of Toronto.
 
SciNet is part of a network of seven Canadian high-performance computing groups.
 
Currently, it is accepting applications from all types of researchers who want to use their facilities.
 
“We’re committed to providing a significant fraction of these resources to people outside the University of Toronto . . .
 
We’re in a sharing mode,” Peltier said.
 
 
 
SciNet projects include:
 
Mappings of climate change models, including how the Earth will be affected by global warming. This research is being done for the Intergovernmental Panel on Climate Change, a United Nations group that shared the Nobel Peace Prize in 2007 with Al Gore.
 
Computing data gathered from the Large Hadron Collider (LHC), a massive particle accelerator in Europe that smashes atoms together at high speed in the search for the Higgs-Boson, or “god” particle. The University of British Columbia will receive the first set of data when the LHC is switched on later this year, then process it and pass it on to SciNet, where they hope to begin getting useful information from the numbers.
 
— source: Richard Peltier, scientific director of SciNet