Skip to content

Architectures For Parallel Computing

Public syllabus for 2025-2026

Academic overview

Programme
AIDC
Period
Year 1, Semester 1
Credits
6
Weeks
14

Teaching team

Course coordinator
(none)
Seminar coordinators
(none)

Learning time distribution

Total
Curriculum Lecture Practice Total Weekly Lecture Practice
42 28 14 3 2 1
Exam hours
10
Individual Study Bibliography study Field study Homework Tutoring Others
108 20 15 58 5 0
Overall
150

Learning outcomes

Knowledge

  • Knowledge of basic architectures and VLSI concept
  • Knowledge and understanding of convolutions and their applications in signal filtering
  • Knowledge and understanding of the theoretical model and applications of long integer arithmetic
  • Understanding challenges and implementation solutions in GPU

Skills

  • Design and implementation/simulation of systolic algorithms in computer arithmetic, matrix computation, numerical computing and graphics
  • Implementation of parallel algorithms that use GPU

Responsibility

  • Ability to use specialized applications in different domains

Online platform

(none)

Course content

Content Methods Obs
L1-2. Introduction, evolution of the model, theoretical models Discourse, conversation,illustration by examples
L3. Parallel algorithms analysis, comparison with sequential algorithms analysis Discourse, conversation,illustration by examples
L4-5. Systolic algorithms for matrices operations Discourse, conversation,illustration by examples
L6-8. Semi-systolic and systolic algorithms for signal processing (FIR, IIR) Discourse, conversation,illustration by examples
L9. Systolic algorithms in multi-precision integer arithmetic Discourse, conversation,illustration by examples
L10-12. Challenges in SIMD architectures and their solution in CUDA, comparison of SIMT with SIMD and SMT Discourse, conversation,illustration by examples
L13-14. Architecture details of GPU's supporting CUDA, optimizations and limitations Discourse, conversation,illustration by examples

Course bibliography

Brudaru, Octav şi Gâlea, Dan – Introducere în CALCULUL SISTOLIC, Ed. Academiei Române, 1994 Jebelean, Tudor – Systolic Multiprecision Arithmetic, Phd. Thesis, RISC Linz, 1994 Knuth, D. – Tratat de programarea calculatoarelor. Vol. 2. Algoritmi seminumerici, Editura Tehnică. Petkov, N. – Systolic Parallel Processing, in Advances in Parallel Computing, Volume 5, North-Holland, 1993 Classroom materials (Classroom code: rdvjsai)

Seminar content

Content Methods Obs
S1. Cellular Automata, Elementary cellular automaton, Game of life Dialogue with students,cooperative learning
S2. Evaluation of polynomials Dialogue with students,cooperative learning
S3. Matrices multiplications Dialogue with students,cooperative learning
S4. Convolutions and linear filters Dialogue with students,cooperative learning
S5. Long integers multiplication Dialogue with students,cooperative learning
S6. Basic operations on images using CUDA Dialogue with students,cooperative learning
S7. Julia fractal computation using CUDA Dialogue with students,cooperative learning
Bibliography: Petkov, N. – Systolic Parallel Processing, in Advances in Parallel Computing, Volume 5, North-Holland, 1993 https://docs.nvidia.com/cuda/cuda-c-programming-guide/index.html

Seminar bibliography

(none)

Corroboration

(none)

AI tools guidance

(none)

Evaluation and delivery

Activity Criteria Methods Percentage
C
  • Dedicated architectures and specialized algorithms
  • Written exam
  • 30.0%
S
  • Lab assignments
  • Oral evaluation. Student presentation. Discussion.
  • 70.0%

Performance standards

Student should have basic understanding of special parallel systems and should be able to give at least two examples of existing systems At least 4 lab assignments are required

Additional info

(none)