Duration: 60 Minutes
Occasion: PhD students physics seminar
Location: Forschungszentrum Jülich
Background of Audience: Physicists, PhD students
Third part of my lecture on CP violation. This time on the menu: CKM matrix, the different parameterizations (the usual with sin / cos; the Wolfenstein one), and the unitarity triangle.
Liked it! Although I, still, severely lack depth in knowledge on the topic, burying into a unknown physics theory is, still, a lot of fun.
I made a GIF animation of the CKM fitter results of the last few years. You find it after the jump.
CKM fitter GIF animation
Duration: 45 Minutes
Occasion: NVIDIA Application Lab Workshop 2014
Location: Jülich Supercomputing Centre
Background of Audience: GPU experts and novices, HEP not-at-all’er
Second talk to report the progress we did in the course of the last year when working together with the Application Lab. The latest results from the paper submission are in (all the while there are still some more latest-er results missing…).
As I had a 45 minutes time slot I brought some exhibitions of our STT detector. I like to do that, to show people we’re not only programming stuff for heating up Silicon, but to actually interact with some real life application.
Official place to find these and other slides.
Duration: 30 Minutes
Occasion: GPU Computing at High Energy Physics, September 2014, GPUs at Low Level Trigger session
Location: Pisa University, Italy
Background of Audience: GPU enthusiasts, HEP physicists. No need to waste time introducing stuff.
Usual talk with some new animations introducing Hough transforms and explaining my current Hough space peak finding status.
After the jump I included the GIFs of the animations in this post. The PDF has static, ›keyframe‹ version of them. During the actual presentation I showed the GIFs.
Occasion: Physics PhD Students Poster Session
Location: Ruhr-University Bochum, Germany
Background of Audience: Physicists, but of different flavors. Some solid state physicists, some plasma physics, some astronomy.
Duration: 25 Minutes
Occasion: Dresden CUDA Center of Excellence‘s Workshop on Fast Data Processing on GPUs
Location: TU Dresden, Germany
Background of Audience: Computer scientists, GPU enthusiasts
As you can guess form the title of the talk, it was exactly the same talk as I held at the GPU Tech Conference. Go over there to look at the slides.
The only thing new: My part for the »Bring your own GPU challenge« session. In the PDF, after the Thank You slide, you find some Hough spaces where I explained my challenge of finding peaks in a parallel fashion. More to that in the week’s review post.
My talk was good, the workshop as well. Unfortunately the last two talks were cancelled and I had to wait more than five ours until my flight would leave :(.
Duration: 25 Minutes
Occasion: Fifth International Workshop for Future Challenges in Tracking and Trigger Concepts 2014
Location: Frankfurt Institute for Advanced Studies, FIAS, Frankfurt, Germany
Background of Audience: Physicists working on tracking — experts!
Usual talk sans tracking introduction, as there was an audience full of tracking experts. People were curious on some of the aspects I presented and the experience we gained. Good!
The other talks were also quite good, maybe a bit too deep-level for me (focusing on vectorization and very specific applications). At the end (and also in between) there was quite a fruitful discussion.
PDF version; Version at Scribd.
Duration: 25 Minutes
Occasion: GPU Technology Conference 2014, Session S4499
Location: Room 212A, McEnery Convention Center, San José, CA, USA
Background of Audience: Computer Scientists
Talk at NVIDIA’s conference for GPU-using developers (usually, everything but gaming). Not too big of an audience, unfortunately.
I (intentionally) spend a lot of time introducing the topic, as most of the people in the audience don’t know anything about particle physics and tracking and stuff.
On the GTC page, there’s a recording of the session, unfortunately with the wrong aspect ratio1 – you find a fixed version at the end of this post. The slides accompanying my talk on the NVIDIA side are a cut-down to print version (= less animations).
In the full PDF version provided here you can click on the nearly-hidden links to more in-depth backup slides on some topics (located next to the page number on its left-hand side – e.g. on slide #33.)
PDF Version; Version at Scribd; Stuff at NVIDIA’s GTC session listing.