DidaWiki

• 19/09/2011 Introduction to the course — MPI lesson 1
• 20/09/2011 MPI lesson 2 —
• 21/09/2011 MPI lesson 3 —

• 26-30/09/2011 No lesson and no question time

• 3/10/2011 MPI lesson 4 —
• 4/10/2011 moved
• 5/10/2011 MPI lesson 5 —
• 6/10/2011 moved
• 7/10/2011 MPI lab 1

• 10/10/2011 lesson canceled
• 12/10/2011 MPI lesson 6 — Collective communication, introduction; BCAST, BARRIER, GATHER. Extent and size manipulation for user-derived datatypes; interaction with point-to-point and collectives (recap of the definition, MPI_GET_EXTENT, MPI_CREATE_TYPE_RESIZED, MPI_GET_TRUE_EXTENT).
• 13/10/2011 MPI lab 2
• 14/10/2011

• 17-21/2011 No lesson and no question time

• 24/10/2011 TBB — Introduction to the Thread Building Blocks Library. History, current programming models (parallel patterns and C++ generics, task generation and scheduling over threads), examples of basic parallel patterns (parallel for, reduce), examples of concurrency-aware containers (hashmaps, vectors, queues) and related issues, concurrency control via mutex objects.
• 26/10/2011 MPI lab 3
• 27/10/2011 MPI lab 4
• 28/10/2011 To be confirmed

• 31/10/2011 KDD and Data Mining Algorithms : the K-means algorithm as a parallelization example. MPI parallelization.

• 7/11/2011 MPI lab 5
• 9/11/2011 MPI lab 6
• 10/11/2011 MPI lab 7

• 14/11/2011 Grid to Cloud Computing Survey of the Grid Computing paradigm. Key ideas, research and implementation issues. The concept of middleware; main functions Grid middleware has to provide. Limits of the approach as it was carried on: scalability, lack of a compelling business model. Differences with respect to Cloud Computing: HW Virtualization, consequences on the business models of Providers and users; reversal of the administration burden, everything-as-a-service metaphor : IaaS, PaaS, SaaS levels.
• 16/11/2011 *
• 17/11/2011

• 21/11/2011
• 23/11/2011
• 24/11/2011

• 28/11/2011
• 30/11/2011
• 1/12/2011

…