Using annotations to reduce dynamic optimization time
Chandra Krintz, Brad Calder
Abstract
Open-access reader
Chandra Krintz, Brad Calder
Abstract
Open-access reader
Dynamic compilation and optimization are widely used in heterogenous computing environments, in which an intermediate form of the code is compiled to native code during execution. An important trade off exists between the amount of time spent dynamically optimizing the program and the running time of the program. The time to perform dynamic optimizations can cause significant delays during execution and also prohibit performance gains that result from more complex optimization.
OpenAlex reports 84 citations for this work. Citation counts describe recorded attention and do not establish research quality.
A contribution statement is not available in the OpenAlex record.
Method details are not available in the OpenAlex metadata.
Findings are not separately available in the OpenAlex metadata.
Limitations are not available in the OpenAlex metadata.
Application details are not available in the OpenAlex metadata.
Dynamic compilation and optimization are widely used in heterogenous computing environments, in which an intermediate form of the code is compiled to native code during execution. An important trade off exists between the amount of time spent dynamically optimizing the program and the running time of the program. The time to perform dynamic optimizations can cause significant delays during execution and also prohibit performance gains that result from more complex optimization.
Key concepts: Computer science, Dynamic compilation, Execution time, Program optimization, Code (set theory), Parallel computing, Running time, Dynamic programming