Intel Academic Community at SIGCSE 2011


The Intel Academic Community had a fantastic time engaging with computer science educators at SIGCSE 2011. Catch up on our videos of SIGCSE booth demos and professor interviews! At SIGCSE, the Academic Community shared methodologies of teaching parallel programming to educators-- including game development, robotics, and hands-on labs with the Intel® Manycore Testing Lab-- and introduced the Intel Academic Community MeeGo Program. We also announced the Academic Community's new Intel® Black Belt Software Developer- Professor Dick Brown, from St. Olaf College. Learn more by exploring our site and be sure to join us on Facebook, Twitter, and LinkedIn to stay in touch! Thanks to all for engaging with us- we look forward to seeing you in 2012!

Learn about the Intel Academic Community MeeGo Program




Watch the videos!
Opportunities of MeeGo in the Classroom
Integrating MeeGo into the Classroom
Academic Program for MeeGo at SIGCSE 2011

 

SIGCSE 2011 Photos, Blogs and Videos

Photos
- SIGCSE 2011 Photo Album on Facebook!
- SIGCSE 2011 Photo Booth Tour
Blogs
- SIGCSE 2011 is starting!
- A Customer, You Say? Making Requirements Engineering Fun!
- It's a Snap to find the Scratch to BYOB in class
- Got Hardware? Exploring barriers and breakers to teaching parallelism
- Dr. Susan Rivoire- Introducing Undergraduates to Parallelism
- SIGCSE 2011 with Ed Carr and Josh Adams

ACIA Autobot - ASU, Carl Hayden, and Intel Autobot
Intel Booth Demo

Presenter: Jesus Yuriar, ASU Ira A. Fulton Schools of Engineering


The ASU, Carl Hayden, and Intel Corporation Autobot (ACIA) robot uses the Intel® Atom™ processor and Intel multicore processors. ASU students used Service-Oriented Computing (SOC) to create a development environment supporting Visual Programming Language (VPL). VPL is simple enough for high school students to program autonomous robots and learn computing concepts. The ACIA robot is equipped with four independent motors and eight sonar sensors, each of which is managed by a service, resulting in a complex multithreading application.

The students conducted experiments on a single core processor as well as numerous multicore processors. ASU students developed the software in their senior capstone courses CSE 423/424. In addition, Intel® Threading Building Blocks and the performance data collected from CSE 423 was used in a CSE445 Distributed Software Development course to demonstrate parallelism and concurrency on multicore architecture. Learn more >>

 

Parallel Computing Curriculum Development Using Video Games
Intel Booth Demo

Presenters: Professor Ashish Amresh, ASU Polytechnic
Ryan Anderson, ASU Polytechic

Multiple demos that apply parallel computing concepts inside a video game environment will be presented. Primary audience for these demos are faculty who are interested in introducing parallel computing courses in undergraduate computing curriculum. Faculty can learn how to set up the software and hardware necessary to administer the game environment used for teaching parallel computing.

• Multithreading, synchronization:
3D World demo: a basic 3D scene will be rendered with a moving camera which tests for collisions against many moving models. Students will need to separate the rendering thread from an update thread will be required to improve performance.

• Thread spawning:
Maze demo – A player will be able to place walls, rats, and cheese on a large map. Rats may pass through each other, but not through walls to get to the cheese. Students will be required to dynamically spawn a thread for each rat, synchronize each threads return value to determine the winner.

• Multiple Consumers:
Target Practice demo – Players will be able to lob missiles at targets. Since missiles are slow, there can be multiple missiles in the air at once. Students will be required to have multiple threads detect for collisions between missiles and targets simultaneously.

High Performance Biometric Recognition
Intel Booth Demo

Presenters: Professor Ed Carr, North Carolina A&T State
Joshua Adams, North Carolina A&T State

A biometric recognition system consists of three main components: the sensor, the feature extractor and the identification/comparison module. The sensor takes a sample, such as the image of a face or the fingerprint of a subject, and converts it to a digital format. The feature extractor then gathers meaningful information or features from the digital sample. Once features are extracted, the identification module compares those features to a database of subjects to attempt to “recognize” that subject.

Genetic algorithms require a large number of function evaluations to be successful. It’s not uncommon to run a genetic algorithm for 1,000 or 10,000 function evaluations. By using OpenMP, the evaluations can be separated among the available cores giving a vast increase to performance.

 

The Intel Academic Community
Intel Booth Demo

The Intel Academic Community provides free resources for professors to integrate parallelism into their undergraduate curricula. Resources include free courseware, student access to a 32-core server (the Intel® Manycore Testing Lab), and support through blogs, forums, and social media (Facebook and LinkedIn).

Learn more about the Manycore Testing Lab in our video from the SIGCSE 2011 Technical Symposium Video Exhibition:

Learn how you can engage students with hands-on manycore testing, demonstrate software scaling up to 32 cores/64 threads, and even conduct research to support parallelism and scalability in the classroom. Prepare your students to compete in today’s job market as parallel programming and manycore processors reshape the computing industry by extending your college hardware and software budgets with remote access to a wide range of industry-leading Intel® software tools and resources.

NSF/IEEE-TCPP Curriculum Initiative on Parallel and Distributed Computing – Core Topics for Undergraduates
Special Session on Saturday, March 12 (10:55am-12:55pm)


A working group from IEEE Technical Committee on Parallel Processing (TCPP), National Science Foundation (NSF), and the sister communities, including ACM, will introduce curriculum for computer science (CS) and computer engineering (CE) undergraduates on parallel and distributed computing. The goal of this committee has been to propose a core curriculum for CS/CE undergraduates, with the premise that every such undergraduate should achieve a specified skill level regarding PDC-related topics as a result of required coursework.

The Intel Academic Community is a co-sponsor of this Curriculum Initiative and its Early Adopter program.

The TCPP session at SIGCSE 2011 discussed the following:
1) Introduction to the Curriculum Initiative -10 minutes (Sushil Prasad, Georgia State University and Andrew Lumsdaine, Indiana University)
a) Background Activities and Roadmap
b) Why a Parallel and Distributed Computing Curriculum now?
c) What should every (computer science/engineering) student know about computing?
d) How to Interpret this Curriculum Proposal?
2) Curriculum and Rationale for Architecture Topics – 10 minutes (Chip Weems – University of Massachusetts)
3) Curriculum and Rationale for Programming Topics – 10 minutes (Alan Sussman, University of Maryland)
4) Curriculum and Rationale for Algorithms Topics – 10 minutes (Arnold Rosenberg, Colorado State University)
5) Discussion and Q&A - 35 minutes



Academic Community Sessions from SIGCSE 2011

Intel Academic Community MeeGo Program
Sponsor Session on Thursday, March 10 (1:45-3:00pm)



Download the presentation! Academic_Community_MeeGo_Program_SIGCSE2011.pptx

Presenters: Selwyn You - Intel Corporation, Jukka Heikkilä - University of Jyväskylä, Russel J Clark and Matthew Wolf - Georgia Institute of Technology (Click here for full biographies)

Have you heard the buzz about MeeGo? The newly formed MeeGo Operating System-a combination of Intel’s Moblin and Nokia’s Maemo-provides a unified development environment for developers with worldwide distribution channels to consumers.

Learn how Intel’s Global MeeGo University program can help you integrate application development on netbooks, tablets, and smartphones into your curricula. Professors from the University of Jyväskylä and Georgia Institute of Technology will share insight into classroom modules that meet demand for rich application development on client devices. They will also share student feedback, in addition to how, why, and when you should start introducing MeeGo to your students.

 

Using the Intel® Manycore Testing Lab to Test & Scale Applications
Sponsor Session on Friday, March 11 (1:45-3:00pm)



Download the presentation! Intel_Manycore_Testing_Lab_at_SIGCSE2011.ppt

Expand your students' learning experiences with the Intel® Manycore Testing Lab-a unique, gobal, remote-access facility made available at no charge to members of the Intel Academic Community. The Intel Manycore Testing lab can be used to test, validate, and improve the scalability of classroom labs, homework, and capstone projects. The lab supports both Linux* and Microsoft Windows* with a 32-CPU/64-thread development environment, including up-to-date, essential performance tools to assist professors and their students.

Bob Chesebrough will demonstrate a command-line ray tracer application created by Master's degree student Dave Frogley. The application was tested, scaled, and tuned on the Intel Manycore Testing Lab. Discussions will include scaling methodologies and state-of-the-art software tools. Bob Chesebrough is a senior course architect in Intel's Innovative Software Education department.

 

Using Game Development to Teach Parallelism: Where can I find resources to get started?
Birds of a Feather Session on Thursday, March 10 (5:00-6:10pm)

What resources are available for students to learn parallel programming? Can parallel programming be taught effectively using game codes? Isn’t game programming difficult and parallel programming more so? What advantage is there in teaching parallelism through games development?

This BOF will discuss opportunities for computer science students and their instructors to access pre-coded games codes, such as TickerTape, Destroy the Castle, or Smoke and use them as base to which they add significant performance through parallelism. Audience members will be able to interact with others that have used game development as a vehicle for teaching concepts in parallelism.

Facilitators: Joel Adams (Calvin College), Ryan Anderson (ASU Polytechnic), Bob Chesebrough (Intel), Tom Murphy (Contra Costa College)

 

Technology Available to Educators - How Does Access to Next-Generation Hardware Impact Students?
Birds of a Feather Session on Thursday, March 10 (6:10-7:00pm)

This BOF will explore opportunities for CS students to gain hands-on experience using state of the art hardware and software in the classroom. Options include cloud service providers, university hosted cloud solutions, and the Intel® Manycore Testing Lab-a remotely accessible, 32-core, 64-thread server that teachers and students can access at no charge. Audience members will share ideas and collaborate on actionable solutions that will directly impact their students.

Discussion points will cover: What resources are available for students to learn parallel programming? Can parallelism be taught effectively without state of the art hardware? What are the minimum hardware requirements to teach parallelism? How can teachers provide students with a consistent environment and tool set?

Facilitators: Jennifer Teal (Intel), Paul Steinberg (Intel), Tom Murphy (Contra Costa College)

Parallelism and Concurrency in the Curriculum
Birds of a Feather Session

Multicore, multiprocessor, and clustered platforms have become the standard computing platforms available for executing programs. Thus, Computer Science education increasingly entails teaching programming in parallel environments. This BOF is designed to gather colleagues to exchange ideas and discuss questions such as the following: What fundamental ideas of concurrency and parallelism should every CS graduate know? How should concurrency and parallelism be integrated in the curriculum? At what levels and depths? If broad integration is appropriate, what resources and initiatives are needed to educate the educators? What are some good/proven practices and platforms for teaching this material to undergraduates in these various contexts?

Facilitators: Daniel Ernst (U of Wisconsin - Eau Claire), Jens Mache (Lewis & Clark College), David P. Bunde (Knox College), Matthew Wolf (Georgia Tech; Intel® Black Belt Software Developer for Academia), Richard Brown (St. Olaf College; Intel® Black Belt Software Developer for Academia), Libby Shoop (Macalester College), and Michael Wrinn (Intel)

 

 

 

Categories:
For more complete information about compiler optimizations, see our Optimization Notice.

Comments