adam.oliner.net

I am under the impression that my workstation at IBM, named Turing, is, in fact, a quantum computer. I was led to this conclusion by the observation that doing the same thing twice does not always return the same result. For example, I used Matlab to graph a data set. Looked at the graphs, closed them, then graphed them again. Different graphs. Additionally, there seems to be some program running on this machine that changes my files without my knowledge. I try to cvs commit a file with Eclipse, and it says the file that’s open doesn’t reflect the file system. I have to refresh, update (because the files in the repository have also been suspiciously changed even though no one else has touched them), and then commit before the files are changed again. I have consulted Information Systems and a Red Hat Linux expert. They have no idea what’s going on.

While I am aware of a number of proposed benefits [pdf] of a probabilistic computational model, I assure you that, for most tasks, determinism is very much desired.

I attended a roundtable discussion today, moderated by Bill Pulleyblank, my boss’s boss’s boss and head of Project BlueGene. The committee member in whose work I was most interested was Lisa Su, director of PowerPC and Emerging Products, IBM Microelectronics. She headed projects like the PPC970 (or G5), the GameCube, and the “cell” processor in the upcoming PlayStation 3. She’s also went to MIT for her B.S, M.S., and PhD. I went up and introduced myself afterwards, quickly asking about IBM and Apple and the development of the G5. I told her I could see myself being involved with future IBM/Apple projects and to remember me.
One of the rountable members got up at one point to “take care of some business.” Unfortunately, he neglected to turn off his microphone, so the whole auditorium was treated to the sound of a toilet flushing. Everyone laughed for about 5 minutes, and he was applauded upon his return.

Favorite Quote: “It’s not a CD case, it’s a DVD case!”