Paradyn/HTCondor Week 2013. The fulfillment of dreams.
Which dreams, you may ask?
Leon Lederman dream
One is the dream of Leon Lederman, Ph.D., Nobel Prize in physics 1988, director emeritus of Fermilab in Batavia, Illinois. He describes it in his 1993 book God Particle. If the universe is the answer, what is the question?.While still in his pajama. Lederman meets the Greek encyclopedic philosopher Democritus,
The dream takes place in some times before 1993.
Lederman: We suspect that mass comes from a field.
Democritus: A field?
Lederman: Our theoretical physicists call it Higgs field
Democritus: Higgs? What is a field?
Lederman: A field is represented by a particle we call the Higgs boson.
Democritus: And have you found this Higgs particle in your accelerators?
Lederman: Well, no.
Democritus: So you found it where?
Lederman: We haven't found it yet. It exists only in the collective physicists mind.
Democritus: So why do you believe in it?
Lederman: Because it has to exist... I suggest you come here in Batavia in 1995.
Miron Livny's dream
The concept of HTC (High Throughput Computing) germinated in the work of Professor Livny at University of Wisconsin, Madison in 1983. At the HTCondor Week 2013 they celebrated a great victory. Here is what HTCondor did
Is this Big Data? If you can have $1 for each collision, the one million billions dollars mean 14x the Gross World Product of 72,000 billion dollars in 2012.
This is one important segment of the team of computer scientists that helped making the discovery of Higgs boson aka God Particle.
 |
| HTCondor Team 2013. Seated are Todd Tannenbaum and Miron Livny |
Leon Lederman invited too early Democritus to Batavia. He said it will be 1995. It was not 1995, but 2012, seventeen years later. It took much, much more than the custom made DAQ (Data Acquisition Computer) from 1993 Fermilab, which seems a midget compared to the computer resources OSG using HTC can locate, mobilize and utilize today from all over the world.
How does it feel to attend a Condor Week
This is an event that has soul. We cut a huge cake for Miron Livny's birthday.
The features of the next HTCondor 8 were presented as cooking competition, complete with apron chef, like a TV show.
The amenities at Morgridge Institute were state of the art and its' architecture inspiring.
The idea of Bosco, which starting from HTCondor 7.9.4 + is part of the download, is to make Condor power accessible to any one with a laptop. One of the reason the big media like New York Times did not come yet to interview people from the Center of High Throughput Computing is because HTCondor is not as easy to use as Mac. This will change, fasten your seat belts.
Jason Stowe from Cycle Computing said in Big Science, the applications with high parallel throughput grow much faster the MPI applications.
This decade we have a plethora of new names: Exa-scale, Big data, Software Defined Networks (SDN), Dark data. Spot Pricing, Advanced Computing Infrastructure (ACI), and so on.
HTC and HTCondor will evolve. Miron Livny said:
“... we have identified six key challenge areas that we believe will drive HTC technologies innovation in the next five years.”The success in the discovery of the Higgs particle show how valuable the know how of Fermilab was for advances in computer science and in physics. This collaboration must continue and will continue in a newly created symbiosis.
Evolving resource acquisition models
Hardware complexity
Widely disparate use cases
Data intensive computing
Black-box applications
Scalability
This is where the project dV/dT: Accelerating the Rate of Progress towards Extreme Scale Collaborative Science starts.
And here is man with dream behind the God Particle and who indirectly motivated the whole thing. .
 |
| Leon Lederman, Nobel Prize 1988 |
Immediately after HTCondor Week, I visited OSG at Fermilab. More about it in a future entry.
Additional references:
1. Marco Mambelli blog with notes from Paradyn/HTCondor Week 2013
2. Derek Weitzel Bosco architecture presentation
Comments