This software can execute programs in a distributed and parallel environment. It is designed to optimize program execution by distributing tasks across multiple nodes or processors. This runtime provides a more efficient way of running complex programs.
Dapper (Distributed and Parallel Program Execution Runtime) is a tool designed to overcome the complexities of developing large-scale cloud and grid computing applications. It is rapidly becoming a go-to for scientists and programmers looking to harness the power of distributed computing resources.
In an age where breakthroughs in science require vast amounts of computational power, the cloud or grid can provide access to resources that would otherwise take days to run. However, physical hardware is not enough, and a programmer needs to craft the computation that runs on it. That's where Dapper comes in.
Distributed and Parallel Program Execution Runtime (Dapper) is a tool that bridges the gap between high-level specifications of a program and the low-level mechanisms of distributed and parallel computing. With its dataflow-oriented approach, Dapper enables users to write code locally in Java and execute globally on the cloud or grid.
The user first writes codelets, which are small snippets of code that perform simple tasks but aren't a complete program. Users then specify how these codelets transmit data to each other via edge relations. The resulting directed acyclic dataflow graph is a complete program interpretable by the Dapper server, which, upon being contacted by long-lived worker clients, can coordinate a distributed execution.
By working under the Dapper model, users no longer need to worry about traditionally ad-hoc aspects of managing the cloud or grid. Dapper handles data interconnects, dependencies, error recovery, code distribution, and starting jobs. It provides an entire Java-based toolchain and runtime for framing nearly all coarse-grained distributed computations in a consistent format that allows for rapid deployment and easy conveyance to other researchers.
Dapper is an excellent tool for scientists and programmers looking to harness the power of distributed computing resources, making it a go-to when seeking a more efficient and straightforward way of handling large-scale cloud and grid computing. Its dataflow-oriented approach enables easy management, allows users to write code locally in Java, and execute globally on the cloud or grid.
In an age where breakthroughs in science require vast amounts of computational power, the cloud or grid can provide access to resources that would otherwise take days to run. However, physical hardware is not enough, and a programmer needs to craft the computation that runs on it. That's where Dapper comes in.
Distributed and Parallel Program Execution Runtime (Dapper) is a tool that bridges the gap between high-level specifications of a program and the low-level mechanisms of distributed and parallel computing. With its dataflow-oriented approach, Dapper enables users to write code locally in Java and execute globally on the cloud or grid.
The user first writes codelets, which are small snippets of code that perform simple tasks but aren't a complete program. Users then specify how these codelets transmit data to each other via edge relations. The resulting directed acyclic dataflow graph is a complete program interpretable by the Dapper server, which, upon being contacted by long-lived worker clients, can coordinate a distributed execution.
By working under the Dapper model, users no longer need to worry about traditionally ad-hoc aspects of managing the cloud or grid. Dapper handles data interconnects, dependencies, error recovery, code distribution, and starting jobs. It provides an entire Java-based toolchain and runtime for framing nearly all coarse-grained distributed computations in a consistent format that allows for rapid deployment and easy conveyance to other researchers.
Dapper is an excellent tool for scientists and programmers looking to harness the power of distributed computing resources, making it a go-to when seeking a more efficient and straightforward way of handling large-scale cloud and grid computing. Its dataflow-oriented approach enables easy management, allows users to write code locally in Java, and execute globally on the cloud or grid.
What's New
Version 0.96: N/A