Parallel Computer Architecture

ABOUT THE COURSE:
With the emergence of application domains requiring large data processing as well as faster compute performance, parallel computing has become a critical component. This involves the use of multi-core processors as well as tightly integrated clusters of PCs and workstations. The most exciting development is the convergence of traditionally disparate approaches of shared memory, message passing, and data driven computing onto the same platform.

As the parallel applications communicate with each other using shared data; it is important to correctly manage the shared data within the memory hierarchy to avoid inconsistency. Here, we need to understand the mechanisms used for data access, communication and coordination of work and its correct implementation.

This course will introduce the field of parallel architectures and discuss in-depth shared memory management for parallel architectures. We will also learn about the interconnection topologies and routing methods which are important for connection and communication of the multiple cores executing the parallel applications. The course will conclude with an overview of various parallel architectures and on-going research directions.

INTENDED AUDIENCE: Final year UG as well as PG

PREREQUISITES: Final year undergraduate as well as post-graduate students in the disciplines of CSE, IT, CS, ECE, EEE, etc. Can benefit from this course. A basic knowledge in Computer Organisation is sufficient to take this subject.

INDUSTRY SUPPORT: Intel, Micron, AMD, IBM, NVIDIA, HP, Samsung, etc.