High energy physics experiments, such as ATLAS and CMS at the LHC, are collecting petabytes of data. This data needs to be moved to the computing facilities that are distributed around the world. High-speed networks are currently used to move the data but as the size of the data samples continue to grow, it is critical that we develop faster networks. The current networks used by the LHC are 10 gigabits/second. This project will show that the new network and computing technologies will allow us to increase the network capability by a factor of 10 to 100 gigabits/second.

Background:
The demo will be done at the SuperComputing 2011 conference in Seattle (November), which is the largest conference in computing in the world, attracting 10,000 participants and all major computing companies.

The demo will move data on the 100G ultra-high speed network between the CalTech booth in Seattle to the UVIC computing centre. The network uses the optical network of CANARIE and BCNET from Seattle to Victoria. Ciena and Brocade provide network hardware. The servers at the University of Victoria are provided, in part, by DELL.

The 100G network will be operational for length of the SuperComputing conference and used to transfer high energy physics data between UVIC and the CalTech booth. As an example of the power of the network, we provide a number of examples. The typical full-length move is approximately 10 gigabits; this network can transfer 10 movies per second. According to Wikipedia, the US Library of Congress hold 235 terabytes of data in April 2011; this network would be able to transfer the entire contents in 5 hours. One petabyte of data from ATLAS or CMS could be transferred in 1 day.

For more information, contact: 
Randall Sobie   
Institute of Particle Physics Research Scientist, University of Victoria
Email : (rsobie at uvic.ca)
Telephone :  (250) 721-7733