Parallel Programming and Operating Systems Quiz

Ideal Programming Language and Environment

• No specific programming language is recommended for the course assignments.
• A Unix-based operating system (e.g., Linux, MacOS) is recommended for the labs.
• A recommended MPI (Message Passing Interface) implementation for the course is not specified.

Parallel Programming Concepts

• Data locality is a concept that deals with optimizing data transfer between cache and computational units.
• A common method to handle synchronization in parallel programming is not specified.
• Task parallelism is a form of parallel programming that focuses on breaking down tasks into smaller units that can be executed concurrently.

Books and References

• The book 'OpenMP' by S.Hoffmann and R.Lienhart focuses on OpenMP.
• 'MPI: The Complete Reference' by Marc Snir is considered the definitive guide on MPI.
• Michael Wolfe's book 'Optimizing Supercompilers for Supercomputers' primarily delves into compiler optimization.
• 'Parallel Computer Architecture - A Hardware / Software Approach' by D.E.Culler and J.P.Singh covers parallel computer architecture.
• The book 'Optimizing Compilers for Modern Architectures' by Randy Allen and Ken Kennedy provides detailed information on parallelizing compilers.
• The pocketbook 'OpenMP' by S.Hoffmann and R.Lienhart focuses on OpenMP.

Parallelism and Its Benefits

• The goal of parallelism is to speed up the execution of programs by executing multiple tasks concurrently.
• Applications that benefit the most from parallel processing are computationally intensive and have a large number of independent tasks, such as protein folding simulations.
• Protein folding simulations require approximately 10^18 floating-point operations (Flop) per time step.

Challenges and Limitations of Parallel Processing

• A major problem associated with parallel processing is the difficulty of synchronizing and coordinating the execution of multiple tasks.
• The increasing computational power and complexity of tasks contribute to the need for parallelism in handling tasks efficiently.
• The primary benefit of achieving parallelism in program execution is to reduce the execution time and improve the overall performance of the program.