Teaching - Vittorio Scarano

"Scalable Computing"

PhD Course offered to the PhD in Computer Science (Dottorato di Ricerca in Informatica) of the Università di Salerno
News Next Class
  • July  5th: Lesson #4 and #5 material inserted, Bibliography updated
 week of July 11th

Description:

The course will introduce the problems and the solutions to scalability issues in parallel programming. In particular, we will first emphasize scalability in multicore processors, then we'll tackle functional programming (Erlang) and map-reduce. These three topics (seemingly unconnected) have as common goal (or side-product, in some cases) the efficien parallel computation on increasingly large data.

Logistics

The course will be 10 lessons (2 hours each) approx every 2 weeks, and will be held in the Meeting Room of the Dipartimento di Informatica (Stecca 7, II piano).

Schedule:

  1. Introduction to Parallel Computing and Scalability [April 13th, 2011]
  2. Scalability and Multicore Processors [April 28th, 2011 (9.30)]
  3. Parallelism in Processors [May 12th 2011 (9.30)]
  4. Energy efficiency and Scalability [May 26th 2011 (15.00)]
  5. "Flavors" of Scalability in Manycore Processors [June 27th 2011 (15:00)]
  6. Introduction to Cloud Computing [mid July]
  7. When scalability affects data: Map-Reduce [end July]
  8. Map-Reduce (II) [mid September]
  9. Map-Reduce (III) [end September]
  10. Erlang  [mid July]
The course is open to all the students (not only PhDs).

Annotated bibliographies

I will try to collect and annotate important and relevant publications for each of the the three topics that will be developed into the course.
  1. Design Principles of Multicore Processors (last update: July 5th, 2011)
  2. Erlang and functional programming
  3. Map-reduce