Transcription of Statistical Debugging for Real-World Performance Problems
{{id}} {{{paragraph}}}
Example: confidence
Statistical Debugging for Real-World Performance Problems
Statistical Debugging for Real-World Performance ProblemsLinhai SongShan Lu University of Wisconsin Madison{songlh, and implementation defects that lead to inefficientcomputation widely exist in software. These defects are dif-ficult to avoid and discover. They lead to severe performancedegradation and energy waste during production runs, andare becoming increasingly critical with the meager increaseof single-core hardware Performance and the increasing con-cerns about energy constraints. Effective tools that diagnoseperformance Problems and point out the inefficiency rootcause are sorely state of the art of Performance diagnosis is pre-liminary. Profiling can identify the functions that consumethe most computation resources, but can neither identifythe ones that waste the most resources nor explain detectors can identify specific type of in-efficient computation, but are not suited for diagnosing gen-eral Performance Problems .}
for real-world performance problems, we study how users notice and report performance problems based on 65 real- world user-reported performance problems in five repre-
Loading..
Tags:
Information
Domain:
Source:
Link to this page:
Please notify us if you found a problem with this document:
Related search queries
Databases Model the Real World The Entity- “Data, Databases Model the Real World, Real, World, The Real World: What Good Is, Real world, Real-World Numerical Bug Characteristics, Who Watches the Watchmen?’’: Ideology and, Real World Collage, Real World Leadership, Rendering of Real-World Environments, Rendering of Real World Environments, Real World Performance of Association Rule Algorithms, Real-World Evidence: What is it, World visual statistics and infants’ first-learned