Parallel Systems | Τμήμα Πληροφορικής και Τηλεπικοινωνιών ΕΚΠΑ

Semester:

5th

Course Type:

Elective Specialization courses (ΠΜ-E)

Track:

CS (Computer Science), CΕT (Computer Engineering and Telecoms)

Code:

ΘΠ04

ECTS:

TEACHING HOURS per week

Theory:

Seminar:

Laboratory:

Specializations

Foundations of Computer Science (S1):

Data and Knowledge Management (S2):

Software (S3):

B Βασικό

Hardware and Architecture (S4):

B Βασικό

Communications and Networking (S5):

Signal and Information Processing (S6):

Related Courses

Detailed Description

DIT_PPS_ΘΠ04-ParallelSystems.pdf

Course Content

Covers the spectrum parallel architectures, parallel APIs, parallel application development and evaluation. Presents Flynn Classification and elaboration on MIMD architectures shared (UMA, NUMA, cc-NUMA) and distributed memory. GPUs as HPC. Measures of speed-up, efficiency, cost, Amdahl and Gustafson Laws. Foster design methodology, data and domain parallelism. Parallel programming APIs MPI, openMp, CUDA and hybrid combinations. Parallel Programming techniques for HPC applications. Profiling tools for evaluating and improving performance.

LITERATURE AND STUDY MATERIALS - READING LIST

Basic Book Peter S. Pacheco, Introduction to Parallel Programming (Greek edition)
Supportive Book Drimakopoulos V parallel systems and programming (on-line) http://hdl.handle.net/11419/3209
Notes, presentaions, tutorials on Parallel programming and tools

e-Class link

ECLASS