English | 中文 | Русский | Français
2,555 Posts served
8,264 Conversations started
As some may (not) know, Intel Software College will be in Austin Texas soon at the Supercomputing 08 conference. While the rest of Intel is out extolling the (very real) benefits of Nehalem and other innovations for HPC, we at software college will be doing a couple of very cool things to advance the goal of establishing a curriculum for parallelism in undergraduate education.
Our goal at Supercomputing is to take some practical steps to help make this goal a reality. We'll kick-off our efforts off with a cross industry and academia birds of a feather event titled -you guessed it - There Is No More Sequential Programming. Why Are We Still Teaching It? Participants at his event include a true cross section of industry, including AMD, IBM, Microsoft, NVidia, Sun, folks from Open Source and Linux, as well as professors from large research institutions including UC Berkeley and the University of Illinois, as well as smaller colleges such as Earlham and Contra Costa College.
• Education Program Birds-of-a-Feather “There Is No More Sequential Programming. Why Are We Still Teaching It?”
Monday November 17, 5:30 p.m. – 7:00 p.m.
Room: 10B, Austin Convention CenterPresenters: Professor Wen-Mei Hwu, University of Illinois at Urbana-Champaign; Dr. Steven Parker, NVIDIA Chief Scientist; Dr. Christoph Lameter, Kernel Developer, The Linux Foundation; Professor Charlie Peck, Earlham College; Dr. Michael Wrinn, Senior Architect, Intel Software College
Moderator: Professor Tom Murphy, Contra Costa CollegeThere have been urgent voices since at least 1995 calling for the introduction of parallel programming into the undergraduate curriculum, yet academic institutions are still teaching sequential programming. This is true despite the fact that all major manufacturers have moved to a many core architecture and current generation CPU, GPU or ASIC designs cannot be efficiently programmed without knowledge of parallel programming. What should we do to train engineers and scientists to exploit the modern compute platform? This panel debates this issue and kicks off a working group of academic and industry experts that will develop and recommend a practical means for creating an undergraduate curriculum with parallelism at its core.
In the audience will be many of the other participants in this endeavor, so this is the first place to come to to participate in the discussion and to ensure that your voice is heard.
We'll be carrying this discussion forward with two talks in the Intel theater on the show floor at the Austin Convention Center:
Open Discussions "There Is No More Sequential Programming. Why Are We Still Teaching It?"
Intel Booth Theater, Tuesday, November 18 & Wednesday, November 19
- Tuesday -4pm- Dr. Tim Mattson, Principal Engineer, Intel Corporation+ Dr. Ben Gaster, Research Scientist, AMD + Dr. Bill Buros, IBM Linux Technology Center + Professor Kathy Yelick, UC Berkeley
- Wednesday, 10am - Dr. Michael Wrinn, Senior Parallel Architect, Intel + Dr. Dan Reed, Director of Scalable and Multicore Computing Strategy at Microsoft + Dr. Steve Heller, Research Director of Sun Microsystems Laboratories + Professor Scott Lathrop, UIUC
Of course not everyone can come to Austin so we'll be doing a webcast as well:
Professor Charlie Peck Associate Professor of Computer Science, Earlham College Professor Scott Lathrop, Associate Director for Education,Outreach and Training National Center for Supercomputing Applications (NCSA) University of Illinois at Urbana-Champaign |
Attend Online – Register here
Attend Onsite - Room ML6, Thursday November 20, 11:00 AM CST |
So all in all, pretty cool. For those attending Supercomputing, please come by to at least one of these events and let us know your thoughts on these topics. For those not at Supercomputing, log onto the webcast and join the discussion. It is both testimony to the import of this issue that we are working together on this, as well as a testament to the collective vision of the really great people who make up this industry that we can get together on something that affects us all. Hmm- I won't be fired, I hope, for calling our competitors great ;-)
Next Blog - a peek at some of the questions we are creating for the Birds of a Feather.
| November 15, 2008 10:16 PM PST
Asaf Shelly
| There should still be serial programming, but it is only some of the time - whenever you have a single thread owner for a resource (like Windows NT Kernel works, see one of my posts) |

Charles