The Expanded VLA will be able to see even more distant galaxies than its predecessor.

But this isn’t your average iMac.

In the words of the man overseeing its construction, this $17-million U.S. machine will be the “biggest, baddest supercomputer for doing this thing that’s been built yet.”

And everything about this computer, from the patented technology that makes it so powerful, to the circuit boards that bring it to life, will be a Canadian product and may soon allow astronomers to see deeper into the universe than ever before.

“This is a very high profile instrument,” says Sean Dougherty, group leader at the Dominion Radio Astronomical Observatory (DRAO) in Penticton, B.C., which is a branch of the National Research Council. “This is going to have a Made in Canada stamp on it, sitting in the middle of the U.S. I think it’s fantastic that we could get Canadian engineering put in such a significant instrument in another country. I think it speaks volumes to Canadian capability.”

This supercomputer has been built to make sense of the data being fed to it by the 27 antennae that make up the Very Large Array (VLA), a group of radio telescopes that have marked the landscape west of Socorro, New Mexico, since 1980.

In this capacity, a specialized supercomputer is known as a correlator. And, Dougherty says, Canadians have earned international recognition as a world leader in building these.

When DRAO met VLA

The correlator currently used for the VLA is a genuine vintage instrument – it’s been in service since the array was dedicated in 1980. Despite its age, it has helped astronomers accomplish plenty in its lifespan, says Bill McCutcheon, a professor of astronomy at the University of British Columbia. Still, he added, without an upgrade, the array would likely be put out to pasture, as its usefulness declined with its aging technology.

In 2001, the United States National Science Foundation approved a 10-year project to upgrade the VLA, including its correlator. When the upgrade is completed, the array will be known as the Expanded Very Large Array (EVLA).

The NRC has built these kinds of instruments for as long as telescope arrays have been in existence – at least since the mid-1970s, Dougherty says.

“We’ve become very familiar with building instruments using the latest technologies. In recent years we’ve completed two outstanding systems that were used for other projects, not just our own,” he says. Those projects included contributions to a Japanese satellite project, as well as upgrades to the James Clerk Maxwell telescope on Mauna Kea in Hawaii, which is joint operated by the Canadians, British and Dutch.

The team approached the Americans, proposing to use WIDAR – a new, Canadian-invented technique that promised to make the correlator for the VLA more powerful and flexible than any other currently operating.

“The VLA has been one of the most successful telescopes in the history of the astronomy and this makes it even more useful,” says UBC's McCutcheon.

The correlator represents the third largest contribution to the EVLA in terms of monetary value – the entire upgrade is budgeted to cost $100 million U.S. – but the Canadian team isn’t receiving financial compensation. Science is different from industry, Dougherty explains, and in return for their contribution, the Canadians will receive access to other projects and credits for things like telescope time that would normally be costly to outsiders.

What's next?

After a prototype tested successfully in August 2008, production has now started on the 256 custom circuit boards that will make up the correlator. By a happy accident, Dougherty says, BreconRidge, the company with the manufacturing contract, is based in Kanata, Ont.

A Canadian technician installs racks for the EVLA correlator. The correlator consists of 256 individual boards.

The hardware for the system is to be delivered later this year, and by January 2010, Dougherty says, the National Radio Astronomy Observatory is expecting the Canadian correlator to be complete and ready to take over from the old electronics. Over the next three years, he says, the rest of the system will be brought online.

So what will astronomers see when they turn on a telescope that has been given a $100-million lease on life and uses some of the best technology available today?

Galaxies, hopefully, says Dougherty. And lots of them.

In addition to further understanding the magnetic structure of the universe, researchers are hoping to discover many galaxies that have been hidden to the current VLA because the signals they emit are so specific and constantly shifting that they have been difficult to pinpoint. With eight gigahertz and 16,000 channels at their disposal, Dougherty anticipates that finding the frequencies of these galaxies will no longer be a problem.

“It’d be like trying to find a needle in a haystack [with the old VLA], but with this new system, it’d be dead easy,” he says.

Front page photo courtesy of NRAO/AUI