 |
Parallel & Distributed Computing |
Intellectual property & Copyright
Announcements | Lecturers | Objectives |
Learning outcomes
| Syllabus
| Bibliography
| Lectures: CSP
PCP
NMA Proj
| Assessment
(Previous year (in Pt): 2010/2011)
Last modified:
06 Dez 2011
| departamento de informática |
 |
Work assignment in CSP: please
read the details in the webpage with the lecture material for
CSP.
25-Nov-11
Contact by email: please note that the lecturer's team has lately contacted the UC students through the mailing list "cpd1112@di.uminho.pt", that contains the email addresses of all students that requested an account in the SeARCH cluster; however, this list is still lacking some students, so the missing students please request a SeARCH account asap through the search_admin mailing list. 24-Nov-11
Schedule for the whole semester:
the table with this schedule is available here,
and it will be updated whenever new data is available.
Updated:
08-Nov-11
Advanced Training Course on CUDA:
under the collaboration program between Portugal and the Texas Advanced
Computing Center at University of Texas in Austin (TACC/UTA), a 4h+4h
training course on CUDA programming (for GPU) is being scheduled for
videoconferencing on two consecutive Tuesday afternoon, on 29-Nov-11 and
06-Dez-11, for all students attending this curricular unit.
(08-Nov-11)
Invited talks: as announced on the
opening session, each week till the end of November the lecturer team will
invite a science/engineering researcher from industry or academy to give a
talk on his/her work project and interests, aiming to motivate students with
ideas for projects for the 1st and 2nd term; those that show relevant
success may become dissertation themes to be offered to interested students.
So far, these events had the following activities:
04-Oct-11, 14h-16h: Stéphane Clain (CM/UM), From
modelling to numerical simulation (slides)
11-Oct-11, 14h-16h: Nuno Micaêlo (CQ/UM), Increasing the realism of
molecular dynamics simulations: incorporation of fast electrostatic
calculations using GPU devices (slides)
18-Oct-11, 14h-16h: Carlos Manta Oliveira (Critical Health), Retmaker:
possible topics for hybrid GPU/CPU processing
(slides)
25-Oct-11, 14h-16h: António Onofre (LIP/UM) and Miguel Oliveira (LIP/UC),
Optimization processes in High Energy Physics: challenges in LHC
(slides)
and
Grid Computing
(slides)
08-Nov-11, 14h-16h: Luís Marques (CF/UM), Simulation methods in
Physics: Core concepts
(slides)
15-Nov-11, 14h-16h: Miguel Nóbrega (IPC/UM), Computer Aided
Design and Optimisation of Polymer Processing Tools
(slides)
22-Nov-11, 14h-16h: Eduardo Pereira (CF/UM), Superdiffusion and
Lévy flights
(slides)
29-Nov-11, 11h-13h: Manuel Melle (CCTC/UM), Realistic modelling
of carbon nanotubes , graphene, and other carbon nanomaterials
(slides)
Updated:
06-Dez-11
Presentation slides: a copy of the
slides that were shown to advertise this course are
here. (03-Oct-11)
Additional activities:
beyond lectures and discussion classes, students are supposed
to (i) attend international workshops or similar events in Braga or Porto,
in topics related to Advanced Computing and (ii) work in a research 5-week
internship (min) at the University of Texas, in Austin (all travel and
accommodation expenses to be covered by the institution. For more
information, consult previous year website. (03-Oct-11)
How this UCE will work:
this UCE is organized in 2011/12 into 3 thematic modules
- which will run in parallel during semester 1 before Christmas, and during
the first weeks of semester 2 - and 1 project-oriented module.
Semester 1 will focus on computing systems, parallelism and
efficiency/performance, while semester 2 will have a stronger numerical
component connected to applications in scientific computing, and the main
part on project work. The project work in semester 1 will basic stress
training studies in application performance evaluation, based on real
problems that will be discussed together with the end users. (03-Oct-11)
Official website: the official website for this course will be at http://gec.di.uminho.pt/minf/cpd/, a virtual URL while the academic year lasts; the real URL is http://gec.di.uminho.pt/minf/cpd1112/ ; some material will be repeated at the institutional e-learning platform (https://elearning.uminho.pt/), namely materials with copyright data, since only registered students may have access to this material. (03-Oct-11)
Computing Systems &
Performance
(CSP)
Alberto José Proença
email: aproenca<at>di.uminho.pt
Parallel Computing
Paradigms
(PCP)
João Luís Sobral
email: jls<at>di.uminho.pt
Numerical Methods &
Algorithms (NMA)
Rui Ralha
email: r_ralha<at>math.uminho.pt
PDC main focus lies on
heterogeneous parallel computing systems and the design of efficient
scientific applications.
While these systems are currently available through computer clusters, the trend
is fast moving into single computing devices, which makes this course a
competence training for next generation of computing systems using current
systems as their close emulators.
Parallelism at multiple levels
is thoroughly addressed in this course:
(i) low-level issues at instruction/data/thread levels are covered under the
organization and architecture of computing systems, including multi-core CPU
devices and many-core gpGPU devices, and shared & distributed access to data;
(ii) higher-level issues of thread/data/process levels of parallelism are better
addressed as parallel paradigms, including methods and techniques for
algorithm/program & data design, scheduling, access and distribution.
To target efficient execution of parallel algorithms and code, this course also addresses methods and techniques to evaluate performance of code execution with profilers and performance analysis tools.
A strong requirement in the design of efficient scientific applications also includes the error control, data stability/robustness, and efficient numerical methods and algorithms for heterogeneous parallel computing platforms.
A student that successfully completes PDC shoud be able to demonstrate that he/she acquired the following competences (knowledge and intellectual & practical skills):
****** under construction **********
Algumas destas competências poderão ser mais detalhadas, de acordo com a ênfase dada em cada um dos módulos temáticos:
****** under construction **********
Conteúdos para cada um dos módulos temáticos e para o projecto integrado:
Computer Architecture: A Quantitative Approach, David Patterson and John Hennessy, 5th Ed., Morgan Kaufmann, 2011
Programming Massively Parallel Processors, A Hands-on Approach, David Kirk and Wen-mei Hwu, Morgan Kaufmann, 2010
Parallel Programming in C with MPI and OpenMP, Michael J. Quinn, McGraw-Hill Education, 2003
Designing and Programming Parallel Programs:
Concepts and Tools for Parallel Software Engineering, Ian Foster,
Addison-Wesley, 1995 (free online at
http://www.mcs.anl.gov/~itf/dbpp/text/book.html, or
here)
Parallel Programming in C with MPI and OpenMP, Michael J. Quinn, McGraw-Hill Education, 2003
Matrix Computations, G. Golub, C. F. Van Loan, 3rd. Ed., John Wiley & Sons 1996
Lecture slides and other additional notes will be supplied during the semesters.
****** under construction **********
This publication - including its printing facility - and associated contents - that might include parcial reproductions of external works adequately cited - is exclusively aimed at the students registered at the Multi-core and Many-core Computing course at the Doctoral Program art MAP-i, at Universidade do Minho in 2010/2011, for personal use and e-learning purposes, and they do not aim any lucrative or commercial goal. Any other reproduction, either total or partial, by any support, medium or process, namely electronics, mechanical or photographic, including photocopy, modification of the contents, its public communication, its distribution through renting, without proper consent from the authors, is illegal and may incur in legal procedures.