Brian Cox. Smashing Research at CERN.

---

Collisions Caught on Camera

Cox’s research at the LHC focuses on ATLAS, which is one of several massive digital detectors inside the collider. Its job is to capture images of the collisions—up to several hundred of the nearly 600 million that happen each second. “You can think of the detectors, really, as 100-megapixel cameras,” says Cox.

The sheer quantity of data the LHC will generate—estimated at a petabyte per month—presents an enormous challenge in terms of sorting the digital wheat from the chaff. Banks of computers at CERN will sift through data and flag anything that might be worthwhile, discarding the rest. “There just isn't enough disk space in the world to record all of it,” says Cox. Any potentially interesting images will be recorded and distributed on the Grid, a version of the Internet developed by CERN that makes data available to scientists anywhere in the world. “Then the trick is to take that data and write C++ computer programs to go through it and try to find the interesting things,” he says. “And that's where the creativity comes in.”

If physicists want to look at specific collisions, Cox says, they use proprietary visualization applications to view the data graphically. “If an electron comes flying out of the collision, then you'll see that in the display. And a proton, you'll see that. There are typically many hundreds of new particles produced in these collisions. It's a big mess, basically,” he laughs. “And you're looking for things like, say, the Higgs particle, which we've never seen before. So to do that, you run computer programs over the data, looking for particular characteristics.”

Many Platforms, One Mac

Cox runs Monte Carlo method simulation programs on the Mac, including PYTHIA and HERWIG, as well as a root analysis package. Physicists at Manchester are also able to run the full CERN analysis suite using VMWare Fusion or Parallels Desktop for Mac.

“On the Mac, we can now run multiple operating systems,” Cox says. “This has been a great thing, because increasingly we're using the Mac Pro instead of Linux machines to run the CERN software. We have 10 Mac Pros with the 8-core Xeon processors. We're finding that they're very reliable, and quite a cheap way of getting very powerful computing.” In addition, because so many people at the Manchester lab are using a MacBook, the Mac Pro provides another benefit: “It’s quite nice to have the powerful Mac Pros, because they interface so easily with the MacBook,” he says.

Physicists have many theories about what might result from the LHC experiment, and anticipation is building around the world. “We're still a long way from understanding the universe,” says Cox. “That's why we're doing this, because we know where our understanding stops. And you can't just work it out on a piece of paper. It's just not the way it’s done. You've got to actually look. Science is quite simple, actually—it’s a process of guessing, and then looking to see if your guess was right.”