Exa-Scale Volunteer Computing

Subscribers:
344,000
Published on ● Video Link: https://www.youtube.com/watch?v=NZRnCwQZe38



Duration: 1:07:47
177 views
4


Peta-scale, the Holy Grail of high-performance computing, is now reality. Sustained PetaFLOPS throughput was first achieved not by supercomputers, clusters, grids, or clouds, but by volunteers. This milestone was reached by Folding@home using 40,000 Sony PS3 game consoles, and more recently by BOINC using 600,000 general-purposes computers. But the story doesn't end there. Volunteer computing offers a plausible near-term path to the next multiple of 1000: Exa-scale. This path involves high-performance processors such as GPUs, Cells, and multi-core CPUs, and perhaps the slower but more energy-efficient processors in mobile devices and appliances. Perhaps equally significantly, volunteer computing undermines the established mechanisms by which computational resources are divided among scientists. And, by involving the public in the scientific process, it promotes scientific awareness, interest, and knowledge.




Other Videos By Microsoft Research


2016-09-06The Gridlock Economy: How Too Much Ownership Wrecks Markets, Stops Innovation and Costs Lives [1/9]
2016-09-06Non-spherical microphone arrays for spherical harmonic analysis of 3D spatial sound fields
2016-09-06The Future of the Internet and How to Stop It
2016-09-06Buying In: The Secret Dialogue between What We Buy and Who We Are
2016-09-06Deterministic Encryption: Theory and Applications
2016-09-063D Object Localization and Shape Matching
2016-09-06More Data Less Work: SVM Training in Time Decreasing with Larger Data Sets
2016-09-06Music Technology for Enhancing Creative Expression and Education [1/3]
2016-09-06Real-time Approaches for Dynamic Structure and Electronic Excitations in Condensed Matter
2016-09-06Network Data Streaming - A Computer Scientist's Journey in Signal Processing
2016-09-06Exa-Scale Volunteer Computing
2016-09-06On Evaluating Language Technologies
2016-09-06Designing a Virtual Information Telescope Using Mobile Phones and Social Participation [1/2]
2016-09-06Edge-Preserving Decomposition for Multi-Scale Tone and Detail Manipulation
2016-09-06Brain in a Bottle, Structure and Algorithms
2016-09-06Building Mashups by Example
2016-09-06Analyzing Metabolomics Data for Automated Prediction of Underlying Biological Mechanisms
2016-09-06Examining Bugs to Improve Static Analysis
2016-09-06Towards Programmable Matter
2016-09-06Automated Assume-guarantee Verification
2016-09-06EUCALYPTUS: An Open Source Service Infrastructure for Elastic Computing Research



Tags:
microsoft research