Grid Hype

When CERN's Large Hadron Collider goes online, a high-bandwidth computer network will crunch the numbers. This new network is called the Grid. To do this, CERN has linked itself with research institutions around the world.

This is a genuine technical achievement. But there is currently some misleading hype. Here are some of the spectacular headlines: The Internet's over.. here comes the Grid, Interweb made obsolete and It’s The End Of The Internet As We Know It (And I Feel Fine).

Some of these stories erroneously claim that CERN invented the Internet. (Readers should take that as a red flag. The US Department of Defense came up with the Internet. CERN invented the World-Wide Web. The web is just part of the Internet.) But there are more significant problems with the hype:

First off, the Internet is not going to be obsolete. At best, we can hope for improvements in the Internet. As a journalist myself, I know the next-big-thing story may sound irresistible. But the Internet will continue to grow and modify. It's a little too big and entrenched for outright replacement.

Second, CERN's Grid is built to handle CERN's data. Yes, it's very high bandwidth. But it's not going to replace consumer connectivity right now. Just consider the last mile problem. It's one thing to lay 1,000 miles of fiber between CERN and a university. It's another thing to lay tens of millions of 100 meter fibers to homes. If the Grid can alleviate bottlenecks in traffic, great. But let's not pretend the whole system will be overhauled just yet.

Third, some of the stories talk about downloading movies in seconds and transmitting holograms. Movie distributors might have a problem with instantly downloadable movies. Also, your current monitor probably doesn't support holographic displays. While the Grid's bandwidth may be able to handle all this data, the hype completely ignores the economic and proprietary interests involved.

Still, what CERN is doing is still quite impressive. According to Scientific American:

The nearly 100 million channels of data streaming from each of the two largest detectors would fill 100,000 CDs every second, enough to produce a stack to the moon in six months. So instead of attempting to record it all, the experiments will have what are called trigger and data-acquisition systems, which act like vast spam filters, immediately discarding almost all the information and sending the data from only the most promising-looking 100 events each second to the LHC’s central computing system at CERN, the European laboratory for particle physics and the collider’s home, for archiving and later analysis.

A “farm” of a few thousand computers at CERN will turn the filtered raw data into more compact data sets organized for physicists to comb through. Their analyses will take place on a so-called grid network comprising tens of thousands of PCs at institutes around the world, all connected to a hub of a dozen major centers on three continents that are in turn linked to CERN by dedicated optical cables.

If this functionality can expand to benefit Internet users at large, beautiful! But please be skeptical of the "end of the Internet" stories. As we all know, the Internet is going to end when we are struck by a giant asteroid without warning.