Concurrency and distributed systems

(Back from an unplanned week of ‘working-holidays’ in Greece… had an excellent time)

It’s not news to Computing Science folks that in many cases concurrency and distributed computing face the same problems. Theoreticians may even say that the two domains do not differ from a modelling point of view and for most problems they would be absolutely right of course.

I have been interested in parallel computing and distributed-memory architectures from my undergraduate years, it was the focus of my PhD research work, and it is an area in which I am very much interested still.

I watched few of the MS Research Faculty Summit 2005 webcasts (via Dan Fay). The one on “Concurrency and Complexity” really attracted my interest. It’s great to see that the same ideas on verifiable, declarative, contract-based, and model-based service-oriented computing which have been in my mind lately are also considered for making concurrency easier to program. I know that there has been lots of theoretical work in the area but it is great to see the approaches being investigated for incorporation in products and mainstream runtime systems and development tools.

“Creating the Personal Supercomputer” was also of interest mainly because I realised how Jim‘s and my views on how to build service-oriented HPC systems match those of Microsoft’s HPC division. How cool.

I am sure you’ll find other cool talks as well (agendas for day 1 and day 2).

I just can’t wait to start working for Microsoft. So many cool things and ideas to look forward to. The US Visa process is frustratingly slow even at premium processing 🙁