Podcast
Questions and Answers
What is the primary reason why applications with long execution times are a concern in supercomputing?
What is the primary reason why applications with long execution times are a concern in supercomputing?
- They violate supercomputer usage policies
- They waste computing resources
- They increase the risk of hardware or software failure (correct)
- They cannot utilize checkpointing
What is the primary purpose of checkpointing in supercomputing?
What is the primary purpose of checkpointing in supercomputing?
- To improve load-balancing decisions
- To aid in performance monitoring and analysis
- To provide snapshots of the application at different simulation epochs
- To mitigate the risk of execution failure and associated losses (correct)
What is a characteristic of checkpoint files?
What is a characteristic of checkpoint files?
- They are exclusively used in system-level approaches
- They are typically small in size
- They can be extremely large (correct)
- They are used for debugging purposes only
What is the term used to describe the resumption of application execution from a saved checkpoint?
What is the term used to describe the resumption of application execution from a saved checkpoint?
What is a secondary benefit of checkpointing beyond mitigating the cost of execution failure?
What is a secondary benefit of checkpointing beyond mitigating the cost of execution failure?
What are the two approaches to checkpointing frequently encountered in HPC?
What are the two approaches to checkpointing frequently encountered in HPC?
What is the primary purpose of checkpointing in high performance computing?
What is the primary purpose of checkpointing in high performance computing?
What type of applications typically require very long runtimes on HPC resources?
What type of applications typically require very long runtimes on HPC resources?
How can application checkpoint and restart be performed?
How can application checkpoint and restart be performed?
Why may an application require large runtimes on HPC resources even with many compute resources?
Why may an application require large runtimes on HPC resources even with many compute resources?
What is a benefit of using application-level checkpointing libraries?
What is a benefit of using application-level checkpointing libraries?
What is a characteristic of high performance computing input/output operations?
What is a characteristic of high performance computing input/output operations?
Flashcards are hidden until you start studying