Martin Purschke's Home Page

My name is Martin L. Purschke. I am employed by the Brookhaven National Laboratory and work for the PHENIX experiment at the Relativistic Heavy Ion Collider (RHIC). For a quick overview of the experiment, see here.
Here you can find a brief CV.

What am I doing here?

I joined PHENIX in 1996 and am the Data Acquisition Coordinator of the experiment. PHENIX has currently about a million readout channels. When RHIC is running, we take data at a rate of about 5000-9000 Events/s, where one Event means the data from one particular collision. This translates into a data stream of several 100 MByte/s to disk. For the largest collision system, Au+Au, we have been running at 800MByte/s (we jokingly refer to this as "one data CD per second".)

On the side, I am the current chairman of the of the BNL Cyber Security Advisory Council (CSAC), whose charter is to "formulate, establish and review BNL's security policies, plans and strategy, and to advise the Chief Information Officer on security and other issues."

I work with a group that builds a Positron Emission Tomograph (PET) Scanner for rats, called RatCAP. The idea is that we image the brain of a rat while it is awake and doing whatever a rat is doing. To the best of my knowledge, this "awake animal" aspect is really new. All PET scanners in operation work only for anesthetized animals. Our rats carry the detector like a small ring around their head and can otherwise move about freely. The problem with unconscious rats is that their brains are asleep, and the anesthesia suppresses the processes one is interested in.
The members of the Physics Department on the project are providing support in building and running the detectors, data acquisition (that's me), simulations (me again), and analysis. We are using the data format and analysis framework of the PHENIX experiment, so the project is using a professional and quite modern framework. For a recent high-profile publication, have a look at our paper in Nature.

Presentations of Interest

The CERN Courier ran an article about our then-new high-speed and large-volume data transfers to Japan using GRID tools. We are doing this routinely now, but back in 2005 this was a still new technology. This topic got later picked up by Science Grid.

I gave an invited (non-PHENIX) talk at the "Computing in High-Energy Physics" (CHEP) conference in Interlaken, Switzerland. That conference celebrated the 50th birthday of CERN. Here are the slides of the talk. The organizers also shot a video of my performance on stage.

The next CHEP conference took place in Mumbai (Bombay). Here are the PowerPoint sides of the talk I gave there.

A Roll-Your-own Linux Rescue CD

I made a script and templates which allow you to easily create your own rescue or setup CD. I use this to set up virtually all machines in the PHENIX DAQ system.

A C++ Course

Some time ago I gave two classes about "Programming in C++". One class took place at the State University of New York at Stony Brook, the other here at Brookhaven Lab. If you follow the link, you can view the presentation, download the course, run the examples. Be warned, though, that the course is rather old, and C++ has progressed. Most of the examples should still work. One of these days I'll update the course and add new material to it.

I got a certification from the SANS Institute as a "UNIX Security Administrator". It involved completing a practical assignment, where you develop and describe a solution for a real-world problem. I chose "Securing a Redhat 7.2 Machine on a Home Network". Although Linux versions are much higher these days, many of the points remain valid. If you run a wireless network at home, have broadband internet access, or have multiple computers on your network, you might get some hints how to improve your defenses against the bad guys.

My address at BNL is
Brookhaven National Laboratory
Physics Department Bldg 510C
Upton, NY 11973-5000
Tel (631) 344-5244
Fax (631) 344-3253
Email: or
If you have downloaded my GPG key from a keyserver, you can verify its authenticity with this fingerprint:
A5AE 1BA2 2702 838A 4E39 9AE6 45DA 112E 2302 D676