Course

Unit 1:

Review of Types of Parallelism – Instruction Level Parallelism – Thread Level Parallelism – Limits of ILP – Parallel Processing Architectures – Superscalar – VLIW – Scheduling – Techniques – Static and Dynamic Schemes – SIMD Architectures – GPU.

Unit 2:

Inter-Processor Communication Schemes – Bus-Based – Shared Memory – Distributed Memory and Network on Chips – Performance Analysis – Data Parallelism – GPU and GPGPU Applications – Overview of HPC Platforms

Unit 3:

Introduction to Markovian/Stochastic Models for Heterogeneous Computing – Operating System – Role of Multi-Cores – Case Study for Heterogeneous Architectures – Case Studies of Data-Intensive Application Platforms – FPGA based Platforms.

Course Objectives

• To introduce approaches to parallelism in modern processors.
• To impart background in communication and memory management schemes in
• multicore systems.
• To provide grounding in the fundamentals of heterogeneous computing systems.
• To introduce typical paradigms in heterogeneous systems.

Course Outcomes: At the end of the course, the student should be able to

• CO1: Ability to understand approaches to parallelism in processors.
• CO2: Ability to apply parallel approaches to analyze computing systems.
• CO3: Ability to analyze the impact of communication and memory architectures on performance.
• CO4: Ability to evaluate the design requirements of applications to be implemented in computing platforms

Skills Acquired: provide in depth knowledge on heterogeneous architectures
CO-PO Mapping:

1. Hennesey and Patterson, Computer Architecture: A Quantitative Approach, Sixth Edition , Morgan Kaufmann, 2018.
2. Uchiyama, F. Arakawa, H. Kasahara, T. Nojiri, H. Noda, Y. Tawara, A. Idehara, K. Iwata and H. Shikano, Heterogeneous Multi-core Processor Technologies for Embedded Systems, Springer, 2012.J.
3. Dongarra and A.L. Lastovetsky, High Performance Heterogeneous Computing, Wiley Series, 2009.
4. Abderazek Ben Abdallah, Advanced Multicore Systems-On-Chip: Architecture, On-Chip Network, Design, Springer 2017
5. Journals, White papers and Technical Reports in the area of High-Performance Parallel computing.