TDT6 Heterogeneous and green computing, Year-2010

Students: Angelo Spalluto, Kjetil Jørgensen, Arve Skogvold, Peter Hemmen, Håkon Karsten Eide, Bahram Najafi Uchevler, Jorge Suárez Cabanas, Jarle Erdal Steinsland, Bent Ove Stinessen, Håkon Haugdal Hitland, Yngve Sneen Lindal, Thomas Bergheim.

TDT6 is a specialist 5th year course on Heterogeneous and green computing. It will include parts of those aspects of green computing that is related to multicore software, ie. multicore programming as well as multicore OS such as scheduling and other techniques that might help reduce the power consumption or improve the performance/Watt ratio. The course will focus on having an up to date reading list of high quality, but with relatively few meetings. In the startup meeting at 16/9 we found that OK timeslots are Monday 1115-1200, Tuesday 1515-1600, and Thursday 1015-1200 for those lectures that can be moved.

Reading and lectures list:

  1. 13/9: Cell BE / PS3 programming - experience, Guest lecture by Mujahed Eleyat (Miriam AS, Halden)) Slides
  2. OMPSs: Efficient and Productive Programming for Heterogeneous Computing, Guest Lecture by Dr. Xavier Martorell 11/10 Paper made available to the students through ITSL. slides.
  3. The use of GPUs for High-Performance Computing, PhD trial lecture by Magnus Jahre 12/10, Slides: show, pdf.
  4. State-of-the-art in heterogeneous computing. Guest lecture by Trond Runar Hagen, SINTEF-ICT. Monday 15/11 slides
  5. Introduction to the Cell multiprocessor, by J. A. Kahle, IBM Journal of research and development, Sept. 2005. (Available to students through course intranet (ITSL))
  6. CellSs: making it easier to program the cell broadband engine processor, J. P. Perez, P. Bellens, R. M. Badia, J. Labarta, September 2007, IBM Journal of Research and Development, Volume 51, Issue 5. 12/10. (Black/white copy available through NTNU-eUBIT, but better version through ITSL)
  7. A dependency-aware task-based programming environment for multi-core architectures by Josep M. Pérez, Rosa M. Badia, Jesús Labarta: . CLUSTER 2008: 142-151. Self-study or colloqium. (Also available through course intranet (ITSL))
  8. Introduction to parallel complexity theory, L. Natvig, section 2.1 in Evaluating Parallel Algorithms: Theoretical and Practical Aspects, Dr.Ing. thesis, 1991, Department of Computer Systems and Telematics (IDT), The Norwegian Institute of Technology (NTH), Pages 13-45. Lectured by Lasse: slides.
  9. Decomposable and Responsive Power Models for Multi-core Processors using Performance Counters, ICS'10, Ramon Bertran, June 2010. (Paper available through course intranet (ITSL))
  10. Energy-Efficient Algorithms, Susanne Albers, CACM May 2010, pp. 86-96. (Paper available through course intranet (ITSL))

Questions can be addressed to Lasse Natvig, preferably by E-mail.

Previous editions: 2009-edition, 2008-edition

2011/05/13 08:54, Magnus Jahre