Oh well, who thought that a move could actually be dangerous?

Well, first, this is the first move in my life where I had to throw away most of my things. Well, we could have taken them to Switzerland but I thought they’re not quite valuable enough to justify the transport costs (~$8000). The only problem in this equation is that my employer would have paid the move but I will now pay for the new acquisitions. Oh well, I guess I’m just nice!

I disposed two full industry-grade (apartment) trashcans like this:

A lot of good stuff :-(. But most of this stuff was still from my student time, so time to get new things!

The apartment was so clean after the move that our landlord even gave us $5 more than our deposit with the comment “that is the first apartment I don’t have to clean” :-).

Everything else was packed into a bunch of suitcases (will be fun to check them all). So a lot of nice luggage:

But before leaving the US, I had to go to the Blue Waters Extreme Scaling Workshop and the MPI Forum. Meaning to haul the luggage around a couple of times … but also life-threatening danger.

So the scaling workshop (where I’m at right now) is held in the northwest suburbs (Des Plaines). At the evening of the first day, we decided to walk a bit around to catch some fresh air. It was 9pm-ish. Well, we shouldn’t have gone into that sided road where a group of about 10-15 people were walking on the sidewalk. Obviously some kind of hispanic gang (like in the movies, seriously). So while we discussed how we avoid them (there was no other street side), a black SUV drove by, and suddenly (without apparent reason), the gang started to throw stones and other things at the car. Seriously!? Right in front of us (10 meters away). Since they were running after the car (on the street), it solved our problem of avoiding them :-). But still, that was a $4000 damage right there. And if this wasn’t enough, the car was coming back (!?) and got another round of stones (don’t ask) and the gang was running in our direction before they dispersed. I guess we were right in the middle of a gang fight. No shots fired, yet. The car stopped in front of us and we decided to go back (the stone-throwers were gone).

Police was at the scene seconds later … on the way back to the hotel, we heard a shot and a bullet deflection sounds (most likely metal) on the same street. Oh man, don’t go for a walk in the suburbs.

More to come … off to the MPI Forum tomorrow.

Actually, since March 1st (I wish I had more time for blogging and other things) I’m a manager, who would have thought that? I was promoted into this interim position after Bob Fiedler, who held it before, decided to retire from UI (now working for Cray). I will most likely hold this position until I leave to ETH at the end of July and I see it as a very interesting opportunity to gain some important experiences in the five month period.

One part of this job is to manage the Advanced Application and User Support (AuS) group in the Blue Waters project. This is not your usual run-off-the-mill user support group but 11 domain experts at masters and Ph.D. level who can talk to the application developers and users as peers. Each so called “point of contact” (PoC) is an expert in a particular domain (e.g., CFD, quantum chemistry, MD, QCD, …) and can advise users at a very high domain-specific level. This is possible because Blue Waters has, as a national resource, a very small user community (approx. 30 teams) and each PoC is working with 2-3 teams and ensures that the system is used efficiently and effectively. It was already very interesting to get AuS up and running for the first early users mid March. We had full support from day one and it went rather smooth!

I am also responsible for application and benchmark performance and certifying SoW items. This sounds much less exciting than it actually is, well, most of it at least :-). It’s a big and fascinating system and network!

Even though my research output may be hurt for some months (I hope my students keep it up ;-)), I think this is a great opportunity for gaining experience in managing a team of professionals and accepting a large-scale supercomputer. Thanks for your trust NCSA!

And even better, I got a new and awesome office! IMHO, one of the best offices in the building:

A large whiteboard! Finally 🙂

A premium view! Well, sometimes the sun annoys me a bit ;-).

And a great view on the beautiful Siebel Center!

I received the SIAM Supercomputing Junior Scientist Prize this year! The description of this biannual award (via SIAM) is

The SIAM Activity Group on Supercomputing (SIAG/SC) Junior Scientist Prize, established in 2009, is awarded to an outstanding junior researcher in the field of algorithms research and development for parallel scientific and engineering computing, for distinguished contributions to the field in the three calendar years prior to the year of the award.

I was invited to the biannual SIAM-PP (Parallel Processing) conference to give a plenary award lecture. I talked about an holistic approach for optimization of parallel codes, similarly to optimizations in serial codes. I hope we will soon be able to automate this process for parallel (at least MPI?) codes!

The presentation is archived at https://live.blueskybroadcast.com/bsb/client/CL_DEFAULT.asp?Client=975312&PCAT=4072&CAT=4075 .

I am obviously very happy that I was selected as the 2012 recipient and want to thank everybody who worked with me (and contributed in this way) in the last years. It was an amazing experience to work with Andrew and colleagues in the Open Systems Lab (now CREST), it certainly widened my horizon significantly. Also, Andrew supported me always pursuing crazy and not so crazy ideas and the freedom I had was unseen. I was almost sad when I had to leave for UIUC, but everybody has to go (temporarily in case of OSL :-)) but there was just no way to argue with or resist Marc Snir (if you know him, you know what I am talking about). I am extremely thankful for the many afternoons Marc spent with me talking about parallel computing and providing amazing insights into the background or “structure” of the problems. I am so happy to be able to work with him. And being at UIUC, Bill Gropp provided me a huge amount of guidance advice (professionally as well as personally) and opportunities to collaborate. I also enjoyed (and enjoy now even more!) the guidance of Bill Kramer in terms of management abilities and problem solving. It’s simply an amazing environment!

Here are some impressions from the award:

The talk was well attended even though it was after 7pm :-).

Kamesh Madduri and David Bader presented me the award.

The plaque!

Another plaque!

I just looked at some statistics from 2011 :-).

I completed 670 tasks (that were enough effort to put them on my tasklist), about 2 tasks a day. I received (and read) 20555 emails (after filtering mailinglists and spam!), about 56 emails/day, I sent 5688 emails, about 16 emails/day. Way too many, it’s already consuming a significant fraction of my time! I flew a total of approximately 95.000 miles (~176.000 km), which made me travel at an average speed of 17.4 km/h over the year. I feel bad for the caused carbon footprint but this means that I also spent at least 300 hours (about 12 days) in planes (this is a lower bound estimate). Oh well! I hope 2012 gets quieter :-).

Another WOW from TSA. So I was on my usual travel from San Jose back home. As usual, I refused to go through the millimeter wave detection scanner (also known as “cancer machine”). While being searched and patted down, I saw a full flight crew walking through the metal detector instead of the empty (!!!) scanner (there was no line at all).

They were also neither patted down nor searched. So I asked about that. The agent replied “Flight crews on duty never go through the scanner.”. First, I asked how they know that they’re flight crews (cf. “Catch me if you can”). Well, apparently a uniform and some form of plastic badge is ok (no list or barcode or anything). I didn’t get a reply to my question “why” (especially if there was no line!) :-).

I guess people dressed up in uniforms with some form of airline ID are much more trustworthy than other fellow travelers. Also, this may be some for of admitting that those scanners are affecting people’s health. Well well.

ETH also announced it in English and German.

Even the hpc-ch blog and Inside-IT.ch reported on it.

The ETH Board has appointed the following individuals as professors: Torsten Hoefler […], currently Adjunct Assistant Professor at the University of Illinois in Urbana-Champaign, USA, as Assistant Professor (Tenure Track) of Computational Science. Torsten Hoefler is internationally regarded as one of the leading young scientists in the field of high-performance computing. At the University of Illinois, he is currently involved in the development of one of the world’s most efficient supercomputers. His research interests focus on system design, programming and efficiency analysis. Torsten Hoefler will provide the Department of Computer Science, the research focus “Scientific Computing and Simulation” and the CSCS (Swiss National Supercomputing Centre) with important stimuli. […]

It’s just nice when you login to your machine and see something like:


11:50:20 up 919 days, 19:21, 1 user, load average: 0.04, 0.10, 0.08
[Mon Aug 29 11:50:20] htor@benten:~


One can make some conclusions:

• 919 days no power outage (the server is hanging off a normal outlet)
• there were numerous network outages during this time
• the software raid also went on a rampage and halted the system for half a day (but came back afterwards)
• 919 days no kernel upgrade (I don’t know of any remote exploits for the running kernel, I won’t tell the version though ;-))
• the hardware (including the disks which make a strange sound since about a year) is rather robust (given that I got the machine from the department’s “dumpster”). The HDs will probably die once they stop spinning :-).

HotI was again a huge success this year! Intel’s facilities were very professional and video recording was just amazing (they had a full camera crew of 5 people including one producer on site). I enjoyed it a lot! And I’ll do one more year of TPC Co-Chair in 2012.

In the meantime, there was also another news article about the conference: EETimes. Unfortunately, I’m slightly mis-quoted but it’s not that bad (but who would built a 17-d Torus, not sure where this came from ;-)).

So consider using it! It’s great! Even insideHPC says so: MPI 2.2’s Scalable Process Topologies. 🙂

More press coverage: http://www.hpcwire.com/hpcwire/2011-08-03/silicon_valley_symposium_heats_up_discussion_of_high_performance_interconnects.html 🙂