• John
  • Felde
  • University of Maryland
  • USA

Latest Posts

  • USLHC
  • USLHC
  • USA

  • James
  • Doherty
  • Open University
  • United Kingdom

Latest Posts

  • Andrea
  • Signori
  • Nikhef
  • Netherlands

Latest Posts

  • CERN
  • Geneva
  • Switzerland

Latest Posts

  • Aidan
  • Randle-Conde
  • Université Libre de Bruxelles
  • Belgium

Latest Posts

  • TRIUMF
  • Vancouver, BC
  • Canada

Latest Posts

  • Laura
  • Gladstone
  • MIT
  • USA

Latest Posts

  • Steven
  • Goldfarb
  • University of Michigan

Latest Posts

  • Fermilab
  • Batavia, IL
  • USA

Latest Posts

  • Seth
  • Zenz
  • Imperial College London
  • UK

Latest Posts

  • Nhan
  • Tran
  • Fermilab
  • USA

Latest Posts

  • Alex
  • Millar
  • University of Melbourne
  • Australia

Latest Posts

  • Ken
  • Bloom
  • USLHC
  • USA

Latest Posts

Ken Bloom | USLHC | USA

View Blog | Read Bio

Computing for particle physics in perspective

1985: First Computing in High Energy Physics (CHEP) conference is held in Amsterdam.

1991 or 1992: I encounter the World Wide Web for the first time. There is no graphical browser for it yet, so I am underwhelmed and not sure what it would ever be good for.

1998: CHEP to be held in Chicago. First time I had heard of the conference, and the thought that popped into my head was, “shoot me if I ever go to that.”

2005: I start to work on computing for the CMS experiment at the LHC.

2007: I attend CHEP in Victoria, Canada. No one shot me.

Last week: 19th CHEP held in New York City, and I was there. There were five hundred people registered, all eager to talk about the latest advances and future directions in software and computing for particle and nuclear physics, and also to explore one of the world’s great cities. (As a native of the New York area, I was happy to play tour guide, although I didn’t expect that I’d end up escorting 17 people to Katz’s over the course of four days.) It was a good opportunity to think about the impact that advances in computing have made on physics.

It’s worth looking at the keynote talk by Glen Crawford of the Department of Energy, who described the role of computing as a key enabling technology for our field. Here is a slide of his that I particularly liked:

On the right is what has become the meme (I guess) that we have been using in the US to illustrate how we need the interplay of scientific explorations in three scientific frontiers — energy, intensity, and cosmic — to understand critical problems in particle physics. But I hadn’t previously seen the diagram in the lower left, which shows the required interplay of advanced technologies to achieve these goals. (It certainly hadn’t occurred to me to put computing on the same footing as, say, the LHC accelerator itself.) Glen goes on to describe how particle physics has long been an early adopter of computing technologies, from networks to grids to the World Wide Web (yes, invented by particle physicists). And, in turn, these technologies have been absolutely necessary to handle the huge amounts of data produced by particle-physics experiments that need to be shared among thousands of researchers all over the world.

Other items that caught my attention:

  • In many ways our data management and distribution problems are similar to those of Netflix streaming movies, except that their total data volume is 12 TB and ours is 20000 TB.
  • Long-term data access and preservation is becoming a growing concern. Particle physics experiments are often unique; it’s hard to imagine anyone will do anything like the electron-proton collisions of the now-defunct HERA collider anytime soon, nor the proton-antiproton collisions of the Tevatron. Perhaps some new finding at the LHC will inspire us to go back and look at old data from other accelerators…but will we be able to?
  • Videoconferencing is required for particle physics experiments to get their work done, given that collaborators are spread all over the world. Making sure the systems for this are robust is an important task.
  • Software “engineering” for particle physics has long deserved to be in quotation marks, given our often haphazard policies for designing and releasing software for experiments. But perhaps our model does really serve our purposes well, and is being used elsewhere in commercial computing development.
  • While experiments often start with widely divergent solutions to software and computing problems, they often start converging after a while, suggesting that there are efficiencies that can be found through cooperation.
  • Computers are evolving in such a way that we will see more and more processing cores on a chip that are supported by less and less memory per processor. We’ll need to make greater use of parallelization, and perhaps figure out how to make use of graphical processor units.
  • Rene Brun is about to retire from CERN. It is hard to imagine that anyone has been more influential in the development of software for particle physics in the past forty years. Ultimately experiments communicate their physics ideas through their software, and Rene’s packages have been the lingua franca of the field. (A sometimes idiosyncratic lingua franca, in my opinion, but the basis of everything all the same.) Rene received a standing ovation at the end of the conference.

    2013: Next CHEP to be held in Amsterdam. Having survived this one without undue violence, maybe I’ll go to that one too. It will be interesting to see which predictions of this CHEP will have come true by then!

    Share