The Intel Developer Forum, in San Francisco August 19-21, brings this year a series of talks and workshops of particular interest to the academic community: a chalk talk on research collaboration in parallel computing, a technical session on expressing parallelism, a threading self-paced lab - details of these and more may be found here.
I'm in Los Angeles for the SIGGRAPH conference. I'm here to look for and write about interesting stuff, and help with a blogger event that Intel is hosting tomorrow to introduce some of the engineers behind Larrabee (more on that in a bit). You can check out what Intel has going on at SIGGRAPH this year on this page.
Having just returned from Gamefest (Seattle, WA), I can conclude that some of these developers attended this conference.
In a recent faculty workshop held at Izmir, Turkey, an interesting topic caused debate amongst the attendees during discussion . And I believe it should not be concerned only by Turkish faculty. That is, where is the proper point to start teaching concurrent/parallel programming in institutes. Supporters on one side would start teaching students concurrent/parallel programming as soon as the programming language courses begin, so that the mindset of parallel thinking and design thoughts can be built in the first place.
If you've ever taken a course on algorithms, you will recall that the goal of this course is to explore different techniques to perform certain computations and to be able to evaluate an algorithm's efficiency in a logical and mathematically rigorous way. During my CS studies I took this course and I've taught the algorithms courses while at the University of Southern Mississippi (both a serial [undergraduate] and a parallel version [graduate]).
It was a hesitation of me to write online blogs. According to the old publishing philosophy formed in my past research life - when I write something to publish, it has to be accurate technically and literally. That is, to me, really time consuming, and I always need to evaluate the return of investment in terms of time - what a "lazy" person I were! :-)
ITiCSE, the ACM (and others) sponsored conference on "Innovation and Technology in Computer Science Education", took place earlier this month in Madrid; Intel was among the corporate sponsors. The event is, in some sense, a European version of SIGCSE, but on a more intimate scale, omitting, almost entirely, the industry-exhibitor portion (think textbooks) so prominent at the American event.
Last week I was musing about getting an iPhone but hadn't committed. I had a decent enough Windows Mobile smartphone that wasn't perfect but was sufficient. Then in a fit of iLuck, I was walking past an Apple store last weekend just as a new shipment of phones came in. My inner Geek took control ordered me to pull out the credit card. I had no activation problems, so an hour of clicking on "I Agree To These Terms" later, I was a proud new iPhone daddy.
So you've slaved over the parallelization of your application and you've gotten some measure of speedup. Is it enough? If so, skip to the next blog post. Or, maybe you can give some further SF reading suggestions for Aharon Robbins's daughter.