CASCON is an annual computer science conference hosted by IBM that is held in Toronto, or more precisely Markham, Ontario. It is run by the IBM Centre for Advanced Studies (CAS) in Toronto.(from Wikipedia >:D<) Working at CAS (Center for Advance Studies) during summer as a CAS Fellow, gave me a great opportunity to work in a very professional environment, which includes dealing with students, faculty members and many good experienced employees (mostly holding PhD degree in Computer Science ). i have been working on the new MASS library for new Cell processor which includes elementary function approximation in single and double precision.... there we go THE SWEET FLOATING POINT NUMBER STUFF ! but seriously it was a very good compensations for my unprofessional knowledge regarding this issues. i have to confess that i have all these from my nice supervisor Dr.Christopher Anand that assign me to this position. the CASCON was a good opportunity to present the project our team was working on. the CoConut project presents a new DSL(Domain Specific Language) embedded in Haskell (a Functional Language) that performs a suitable platform for mathematicians with novel ideas. it`s because of simulated Cell`s assembly function embedded in Haskell, so fair enough for those who know nothing about Cell`s assembly language :). we also have our own scheduler in order to produce optimized and efficient cell assembly language and job was to finally make it ready to run on target architecture (Cell BE). anyway!!! lets forget the details and talk about the conference a little bit :D. the conference was located at Sheraton Hotel in HWY7/Leslie in Toronto , quit a well known place, however the hotel was not as prestigious as i thought before. we had a poster in middle-ware section of the technology show case in the conference. each team have a section with a huge poster indicating the project`s name and the author names. so this is what we had on the board: CoConut-Optimizing code generation on Cell Arcitectur - McMaster University
Mehrdad Mozafari
Shiqi Cao
Christopher Anand
Wolfram Kahl
i think they were about 40 to 50 posters in the that area and all were sponsored by IBM so imaging what a huge funding is beside this whole story.
the hotel`s service was good enough . each day they gave breakfast, lunch and a snack at the evening. i met few Persian students mostly from University of Waterloo. There was a workshop for Compiler at the first day. and something weired happend in one of the presentation a guy (aged 25 to 27) was presenting his project and after 10 minute s the audience asked him few questions and he couldn't answer them i mean couldn't satisfy them , so he stopped to talk and then it seemed that he was overstressed in a way that he couldn't continue the talk. i thought he was about to fall down but few people went and help him sit and the talk manager gave a break right at that moment.
( in CASCON writing the blog, with a good feeling )
the workshop are so hard to understand unless you have a clear knowledge about state of the art, few things that draw my attention are:
-nobody tried to self praised
-nobody tried to impose his/her idea upon others, it was just a suggestion or a comment, thats all.
-they were all professionals, more that i imagined.
-at the end of the workshop there was a panel discussion mainly about about vague and unclear aspect of the cell that was going to be worked on in future.
- they were really trying to learn from each other in a humble way...
-being in a big, well known conference make you more motivated on the job you are working on(its because you know lots of more people had a similar point of view as you do)
-beside all the academic atmospheres is was kinda geeky environment, every body was talking about a shitty thing about compiler, algorithm complexity..... lol
-thats it
Mehrdad
