Up Close and Personal

This page has been the most difficult part of writing my personal homepage. How do you present yourself without showing off too much or making a fool of yourself? On the other hand, isn't this what having a personal homepage is all about?



(In case you're wondering, the picture was not taken in Bochum but in 2003 on a beautiful day trip to Venice.)

BIT ROT ALERT: Due to taking a new job at PENN this part of my homepage needs a serious update.

About My Current Job

My current job is officially the position of a 'resaerch assiociate' in the group of Prof. Dominik Marx at the Lehrstuhl für Theoretische Chemie of the Ruhr-Universität Bochum . This is the kind of position usually offered to postdocs while they are struggling to make a name for themselves and figure out a way to get a real position, e.g. as a professor or something equivalent. You have to help with running the department, do some teaching and in exchange you get (a bit of) money and you may use university and departmental facilities for your research to further the progress of science.

In my case, the situation it is a little bit different. For starters, i don't plan to become a professor ;-). Furthermore, i don't have to do any research (to my own annoyance, i only get to do research if i neglect my other duties or work day and night). Instead my main task is to plan, acquire, organize, configure and run the quite substantial (as of spring 2004: >150 nodes, about 1 TeraFLOP/s, about 200GB RAM, about 10TB disk) computer facilities of our department. Of course, this can be only done in close cooperation and with the help of my colleagues, and although the computing center provides the bulk of our network infrastructure, we have many liberties and as such are quite independent and thus very flexible. The high point (so far) was the purchase of our current main local computer resource, a 76-node dual AMD-Athlon Cluster with a high-speed interconnect.

Since i started here in Bochum (spring 2000), i had the nice opportunity to build a full fledged 'computing center' basically from scratch, with desktop workstations, SMP compute servers, smaller and large parallel clusters, mail server, firewall, web/ftp-server and so on. Also, due to the need of maximizing the cpu-power yield, we are running mostly (and very successfully) linux on PC hardware. This needs constant (re-)evaluation of available hardware and a careful search for optimal and reliable hard- and software combinations, and - since the amount of manpower for maintainance for all those machines is very much limited - many administrative tasks had to be (partially) automated (i.e. a ton of customized, robust and simple programs or scripts had to be written). So in a way you could call me a righteous BOFH and occasionally you may even find me in search for a big enough clue-by-four.

Another facet of my current job is to help adapting, extending and optimizing software for our platforms, mainly our bread-and-butter application CPMD (for which i also provide some user support over the CPMD-mailing list and am currently trying to assemble an on-line tutorial), and to generally figure out stuff. Some results of that are visible on the web, e.g. the framework for building our (partially) multi-language department homepage with a consistent design and automatic pdf file generation (without the overhead of a content management system), or a tutorial for creating high quality visualizations from (ab initio) molecular dynamics or quantum chemical data, or a simple but efficient book search engine.

To fulfil my teaching duties (see above), i have been made the formal head and one of the chief architects of our 'Virtual Laboratory'. Virtual in this case does not mean it is a non-existent laboratory, but it stands for doing chemistry with a computer instead of real chemicals (thank god) and using electron structure methods and molecular dynamics simulations. In other words it is the place where we do practical exercises with our students. Similar to the computer equipment, the virtual laboratory had to be build from scratch and is now a room full of Linux PCs and collection of exercises on our web-server (sorry, not (yet) avaiable from the outside). Although the bulk of the exercises are in german a sample exercise has been made available and translated to english.

Working in and for the 'Virtual Lab' is another high point of my current job (and a quite successful one, too). There is nothing more rewarding than helping students to finally put their (painfully) acquired theoretical knowledge to some use. It also helps a lot to keep yourself in perspective with the non-scientific world, since students (sometimes even rightfully so) tend to question many things you have accepted as normality and use or do without thinking about it.

About My Previous Jobs

Before accepting the job here in Bochum, i had a few other jobs. Right after finishing school, i worked in an industrial bakery mainly manufacturing (cheap) sponge cakes and (reasonable) 'Christstollen'. Not the nicest place to work in, but you could earn an honest buck and you got respected if you worked hard. Not to forget, people always tell you, how good you smell, if you work in a bakery (as long as you did not spend most of your shift cleaning out the garbage containers, that is). After that i did my 20 months of civilian alternative service (the kind of service you have to do if are a concientious objector and refuse to be available for military service) in a youth community center.

After that i had another stint in an industrial bakery, albeit a much better one (i still buy their products), where they actually cared so much about the quality of their product, that they not only refused to buy cheap, inferior ingredients, but they also cared very much about their workers being able to manage their workload properly, i.e. they preferred to have one worker too many to having one to few and a shitty product. Quite a unique experience so far.

Finally, still as an undergraduate studying chemistry in Ulm, i started working as a scientific helper, performing mathematical exercises and computer courses for novice students. Later i 'moved up' to doing exercises for undergraduate theoretical chemistry students, and during my time as a PhD student i was hired as an instructor overseeing the chemistry lab course and exercises for medical students.

At the same time i got initiated to UNIX(tm) system administration, soon co-administrating the various AIX and IRIX workstations of the faculty, that were run by the department of theoretical chemistry. With the advent of the Intel Pentium cpu and due to the tight budget of the department, we figured, that running a bunch of PCs with Linux, would give a better value for the money, that regular workstations. Since linux distributions at that time (there weren't that many, and only Slackware has survived) had a lot of rough edges, and were quite a lot behind the (fast moving) cutting edge, i ended up replacing all precompiled binaries and scripts (including c-library, compilers and boot-up scripts) with hand-compiled, consistently configured versions that were optimized to our local needs. Some of the machines running the final version of that 'custom linux distribution' are still running very well (and you cannot run any current binary there, since they are still running on libc-5 and a 2.0 kernel).

The high point of that time was the assembly of our own little parallel (beowulf) computer. At that time (1997) large parts of the university network were still running on 10Mbit (BNC cable) ethernet and 100Mbit switches were extremely expensive. So we sat down and figured, that we would get a similar (if not better) performance, if we used a lot of 100Mbit ethernet cards and connected them with crossover twisted pair cables in a hypercube topology. Sorting out the routing tables was weird kind of fun (took me a few days), but finally we had a setup (click on the image to see a larger scale version) where we could even control the assignment of individual parallel tasks to certain cpus via IP numbers. As it turns out, my involvment in system administrator activities and intimate knowledge of the inner workings of a linux system, proved to be a perfect preparation for my current job.

Altogether i would say, the common elements of my professional career, so far, are the pleasure i draw from teaching practical courses and my affinity to technology and computers in particular, both bridged by a deep urge to figure out complex interactions and get to the root of a problem. Working at a university in the field of computer simulations of chemical problems (i.e. theoretical chemistry) seems to provide all that (at the price of having to deal with a rather sluggish administration and not much money in your pocket).

Other Interests

My first concious encounter with something resembling a computer was around 1979 when my physics teacher loaned me his programmable Texas Instruments Ti51-III scientific calculator. Due to its severe memory limitations (50 keypresses) i soon had to learn, how to optimize 'programs' by doing the same work with less keypresses. A few years later i got my first own programmable calculator, a Sharp EL-5100. A very unique calculator, since it had an alphanumeric dot-matrix LCD display for 80 characters. Also you could type mathematical expressions like you wrote them on paper and it would only execute it, when you hit the '=' key. In case of an error you could even recall the full expression and correct it. On top of that it could store custom expressions with an argument to be executed like the regular preprogrammed functions. It also had a statistics mode, where you could calculate averages, standard deviations and even linear regressions. After the EL-5100 ceased to function, i owned a Casio FX-602P. This one could be connected to an audio tape recorder where you could save your programs to cassette tapes (sounded quite like a modem). It had a very flexible and memory efficient programming language where you could even program little games. It also could be programmed to play little tunes. If you are interested in the history of (programmable) calculatior there are some terrific resources with lots of images and descriptions on the web, e.g. http://www.rskey.org/, http://ernst.mulder.com/calculators/, or http://www.calculators.de/. If you want to find out what kind of chip is driving your calculator, try http://forensics.calcinfo.com/.

The final switch to real computers happened, when my brother agreed to share his Commodore C64 with me. A computer which was loved and hated at the same time. Although it had a very efficient CPU (~1MHz), high quality sound and graphics (compared to other machines of that time), but most of those features were ony partially available (via PEEK/POKE) from the standard BASIC interface, which was horribly slow to boot (heavy C64 users will still remember the dreaded 'Garbage Collection' that could freeze the machine for hours). On top of that, about half of the 64kB RAM was not available in BASIC mode. But if you learned a bit of assembly programming, suddenly the clumsy box turned into a very capable computer, and soon enough people were programming lots of little helper programs that overcame many of the limitations. The had a very strong following and spirit of C64 still lives (although mainly in emulators), see e.g. http://www.c64.com/ or http://www.c64.org/.

Although i now own and use a pretty up-to-date x86-compatible personal computer my fascination for classic computer hardware continues through my obsession for the classic so-called 'IBM model M' keyboard. Truely one of the best keyboards every made. Built to survive a nuclear war and equally well suited for heavy typing and close combat. I currently own three of them, and one of them has been in heavy use as my workplace keyboard now for over 9 years (vintage 1991, so already 3 years old, when i started using it) without showing signs of use. You supposedly can still buy them at http://www.pckeyboard.com/ But using an original, more than ten years old piece of computer equipment, that has not lost any of its usability during its existence, has a special kick. And i found, that i am not alone with this obsession. There even seems to be some hope to extend its live into the age of 'legacy free' computing.

I occasionally play computer games, but the current batch of shoot-em-all-up-3d games does not interest me much. I rather prefer games where the graphics don't matter that much, because they have a real story, e.g. the classic LucasArts Games: 'Zak McCracken', 'The Secret of Monkey Island', 'Monkey Island 2: LeChuck's Revenge', 'Day of the Tentacle', 'Sam and Max Hit the Road', and 'Grim Fandango'. I used to replay them once in a while until they became incompatible with my current hardware. Thanks to the people at http://www.scummvm.org/ i can now replay them again.

This part of my homepage is still work in progress, please try again later.

back to top

---