Since 1987 - Covering the Fastest Computers in the World and the People Who Run Them

July 23, 2009

Big Ideas in HPC

Michael Feldman

When you cover the high performance computing community as I do, it can be easy to get lost inside our little corner of the world. This is especially true if you just talk to HPC vendors all day. Of course, the whole reason to use cutting-edge computing in the first place is to solve big problems in the real world. And some of these represent the most interesting challenges of the day: climate change, genomics, the nature of the universe, and artificial intelligence, to name just a few.

I was reminded of the last topic because of presentation this week by Henry Makram, director of the Blue Brain Project, based in Lausanne, Switzerland. The project’s goal is to simulate mammalian-style brains in silico. The researchers just happen to use a 10,000-processor IBM Blue Gene supercomputer to get the job done. To date, they have been able to simulate about 50,000 neurons of a rat’s neocortical column in something approaching real time.

At the TED Global conference in Oxford, England, on Wednesday, Makram predicted that within 10 years they’ll be able to simulate a human brain (presumably with a much more powerful computer than the current Blue Gene). In principle, if the model is accurate, the artificial brain should respond like a real human, or at least like Dick Cheney. According to a BBC report, Makram quipped “And if we do succeed, we will send a hologram to TED to talk.”

I bring this up not so much to spotlight the Blue Brain work, which is certainly fascinating in its own right, but to point to the rarity of HPC visibility in venues like the TED conference. TED (which stands for Technology, Entertainment, Design) is quite an interesting organization. It’s a non-profit, established in 1996 by Chris Anderson, editor-in-chief of Wired magazine and author of The Long Tail. The organization is all about the confluence of technology, science, art, and culture. Its catch-phrase is “Ideas Worth Spreading.”

Besides hosting conferences, TED maintains a Web site that hosts blogs and video presentations of movers and shakers in the government, arts and sciences. The content manages to be intellectual and fun at the same time. It’s certainly not a melting pot of content like you find on YouTube. Another Web site along the same lines, although somewhat newer, is Big Think. I peruse this one from time to time and always find something worth watching.

What’s hard to find at TED or Big Think are the HPC visionaries. This may be because most of the top-end academicians in high performance computing are used to attending the same ACM and IEEE sponsored events every year. At the other end are the vendors, which go mostly to trade shows. HPC users, like Makram, may venture further afield, but are usually focused on talking about their applications rather than the wonders of supercomputing. That’s understandable.

I don’t want to leave you with the incorrect impression that our community is totally invisible. Besides the Blue Brain presentation at this week’s conference, I also came across an interesting TED video of something called the AlloSphere. It’s a 3D immersive theater at the University of California, Santa Barbara that uses visualization and audio to present complex data. Rather than trying to describe it further, I’ve inserted the 7-minute video below. If you want more information about the project and the woman behind it, follow this link.



Share This