Michael Wrinn: Intel Manager - challenges in parallel computing, insights on disruptive innovation

This week, Stephen Ibaraki has an exclusive interview with Michael Wrinn.

Michael Wrinn manages Intel's Innovative Software Education team, which collaborates with universities to bring parallel computing to the mainstream of undergraduate education. He also works with the ACM Education Council to bring industrial perspective to curriculum evolution. Prior Intel roles include managing Intel's software engineering lab in Shanghai, and directing research on human interface technology. He was Intel's representative for the original OpenMP specification, and remains active in the parallel computing community. Before joining Intel, Michael worked at Accelrys, implementing commercial and research simulation codes on a wide variety of parallel/HPC systems. He holds a B.Sc. and Ph.D. (in quantum mechanics) from McGill University.

To listen to the interview, click on this MP3 file link

The latest blog on the interview can be found in the IT Managers Connection (IMC) forum where you can provide your comments in an interactive dialogue.
http://blogs.technet.com/cdnitmanagers/

DISCUSSION:

Interview Time Index (MM:SS) and Topic
:00:47:		Michael shares when and how he first got interested in computing.
:02:36:		Can you describe your work prior to joining Intel? "....Straight out of school I got a job with a company, Accelrys, a company developing modeling software....We were sort of joined-at-the-hip with top researchers and at my end in the company our mission was to release robust code for users to run their modeling. Because of the state of the art at the time, we were pretty much forced to be directly involved in the research ourselves. It was fun...."
:03:39:		Can you profile your prior work at Intel? "....I was hired initially to join an effort more into technical computing at that time in the late 90's...."
:04:28:		What are your current research interests and roles? "....At the moment our mission is to bring parallel computing into universities....Once again I've stumbled into a kind of fun combination, where you are working with the source of the knowledge and at the same time with the people who are going to utilize that knowledge...."
:06:03:		What kind of platforms do you offer to educational institutions so that they can investigate parallel computing? "....In order to let university professors get familiar with these challenges and globally develop a portfolio of good examples....we actually made a many-core testing lab available to academics to come and play...."
:08:05:		The shift in computing hardware to parallel systems is well underway. Sequential chips are no longer designed and the era of von Neumann architecture passes into history. How can we be best prepared, and for what, exactly? "....There is, in principle, an awful lot of computing power coming online with the increase in the core count, successive generations and we'll see more of that sort of scaling going forward. But it means that the people who write software will have to get a lot savvier about how to do that. That's a huge challenge but I think also a great opportunity...."
:10:09:		Describe the major parallel programming models. "....There are multi-core eventually leading to many-core systems, and shared memory systems with the chips there in the box as compared to clusters. In the shared memory systems the dominant approach is threading whether it be in a C-derived language or Java or a Dot.Net managed type language, where the implementation is all done through a threading model. The other one, which is pretty much the norm in high performance computing, is a message pass model and that was foisted upon that community for the nature of the platforms they work on...."
:12:42:		What can you say about the time lags between research findings and classroom experience? "....I would say that computer science has traditionally been one of those fields where there has been a significant lag between the cutting edge research and what shows up in undergraduate training. One of the things we (my team and project at Intel) hope is to accelerate and to shorten that gap in time...."
:15:34:		What are some of the university/industry collaborations now underway? "....There's a joint sponsorship between Intel and Microsoft currently at two universities, UC Berkley and University of Illinois Champaign....The hope there is that out of some of that research comes directions to go...As soon as anything shows up in either of these joint lab projects at Berkley or Illinois that looks good for teaching, we'll just grab it and try to distribute it out to all our academic communities...."
:17:00:		You recently spoke at a conference (this involved parallel computing) - can you tell us more about this conference and the outcomes? "....The conference is the ACM SIGCSE (The specialist group conference in Computers and Science Education). It turns out to be the primary one for that community....I used that slot to talk about parallel computing and I took a couple of approaches to try to jolt the audience...."
:20:29:		Can you describe your roles with non-profits and your goals? "....Non-profits being universities....What they will get at least is classroom teaching licences to any of the Intel tools. They'll get access to all the material developed by that community by us and any of our joint efforts with the research labs....We equip the people in this community the best we can - sometimes with micrograms for curriculum development, sometimes we will arrange workshops for them and we have the many-core testing lab where members of this community can log in and try their hand at 32A homework assignments...."
:22:35:		You have volunteer activities with non-profit societies, for example, Association for Computing Machinery (ACM). Can you describe some of your work there? "....I work with the Education Council where I bring an industry point-of-view to the curriculum development....The ACM Education Council and ultimately the ACM Education Board periodically produces a set of curriculum guidelines...."
:24:01:		Michael shares an inspiring story from his experiences judging at the International Science and Engineering Fair (ISEF). "....Unfortunately I didn't get to do it this year but I really do try to get there because it's always inspiring. I judge in the physics category...."
:31:05:		Which disruptive experiences most shaped your life and work? "....I'm like those students at the science fair who learned programming to solve a different problem, which is how I got into parallel computing....It was a gradual thing, skills acquired over years and then at some point I pursued primarily those instead of the sciences...."
:32:90:		Past, present, and future - name someone who inspires you and why is this so? "....Richard Feynman...."
:33:19:		Are there any challenges that you were not able to overcome? "....Everyone has aspirations that usually exceed their achievement and I've got plenty of those...."
:34:08:		How would you describe your top innovative achievement? "....I was a minor player but it was fun to have my own wish list that I had already logged, implemented and realized in a standard...."
:35:40:		Which resources would you recommend? "....The resources for parallel computing: Intel (aggregate of information): http://software.intel.com/en-us/academic/ ACM portal: http://portal.acm.org/portal.cfm Ensemble: http://www.computingportal.org/home...."
:38:04:		Provide your technology roadmap for the future within the confines of what we've been discussing? "....Multi-core chips....many-core chips....I think the future will be heterogeneous, many-core processors and that will present even more interesting programming...."
:41:58:		You choose the topic area. What do you see as the top challenge facing us today and how do you propose they be solved? "....'What are the challenges in computer science education?' As it turns out, there's a group of us writing an article and it will be published presently in communications with ACM on exactly this topic...."
:44:46:		Please provide your views on the IFIP IP3 program on professionalizing the profession. [http://www.IPThree.org] "....One question I would pose is: Given that the practice of computing goes well beyond traditional IT, are they looking ahead to the adjacent areas? I certainly applaud the intent of what they are doing and I wish them well..."