P . METRIC : Tracking Memory Bottlenecks via Binary
نویسنده
چکیده
MARATHE, JAYDEEP P. METRIC: Tracking Memory Bottlenecks via Binary Rewriting (Under the direction of Assistant Professor Dr. Frank Mueller). Over recent decades, computing speeds have grown much faster than memory access speeds. This differential rate of improvement between processor speeds and memory speeds has led to an ever-increasing processor-memory gap. Overall computing speeds for for most applications are now dominated by the cost of their memory references. Furthermore, memory access costs will grow increasingly dominant as the processor-memory gap widens. In this scenario, characterizing and quantifying application program memory usage to isolate, identify and eliminate memory access bottlenecks will have significant impact on overall application computing performance. This thesis presentsMETRIC, an environment for determining memory access inefficiencies by examining access traces. This thesis makes three primary contributions. First, we present methods to extract partial access traces from running applications by observing their memory behavior via dynamic binary rewriting. Second, we present a methodology to represent partial access traces in constant space for regular references through a novel technique for online compression of reference streams. Third, we employ offline cache simulation to derive indications about memory performance bottlenecks from partial access traces. By examining summarized and by-reference metrics as well as cache evictor information, we can pinpoint the sources of performance problems. We perform validation experiments of the framework with respect to accuracy, compression and execution overheads for several benchmarks. Finally, we demonstrate the ability to derive opportunities for optimizations and assess their benefits in several case studies, resulting in up to 40% lower miss ratios. METRIC: Tracking Memory Bottlenecks via Binary Rewriting
منابع مشابه
P58: Visual Working Memory Performance Based on Saccades in Children with and without Specific Learning Disorder: An Eye-Tracking Study
Some of the previous studies show that children with SLD have deficits in visual processing and working memory. Hence, the aim of this research was to investigate problems of visual working memory based on behavioral neuroscience method, using an eye tracker device. The method of present study was ex-post facto study. The participants included couple of twelve children with SLD (mean age=10.92)...
متن کاملAdaptive Multiple Component Metric Learning for Robust Visual Tracking
In this paper, we present a new robust visual tracking approach that incorporates an adaptive metric learning in a multiple components framework. Using a similar overall approach to other state-of-the-art tracking methods, which pose object tracking as a binary classification problem, we firstly employ a new feature selection mechanism based on adaptive metric learning for constructing a discri...
متن کاملAnalyzing Memory Performance Bottlenecks in OpenMP Programs on SMP Architectures using ccSIM
NAGARAJAN, ANITA Analyzing Memory Performance Bottlenecks in OpenMP Programs on SMP Architectures using ccSIM. (Under the direction of Assistant Professor Frank Mueller). As computing demands increase, performance analysis of application behavior has become a widely researched topic. In order to obtain optimal application performance, an understanding of the interaction between hardware and sof...
متن کاملLooking at Nothing Indicates Memory Search in Multiattribute Decision Making
The common methods for studying heuristics in memorybased multiattribute decisions provide outcome and response time data but leave the foregoing cognitive processes in the dark. We demonstrate a novel process-tracing method that uses the looking-at-nothing phenomenon to study memory search and cue processing via eye tracking. Participants learned cue information of decision alternatives in spa...
متن کاملMetric Space Searching Based on Random Bisectors and Binary Fingerprints
We present a novel index for approximate searching in metric spaces based on random bisectors and binary fingerprints. The aim is to deal with scenarios where the main memory available is small. The method was tested on synthetic and real-world metric spaces. Our results show that our scheme outperforms the standard permutant-based index in scenarios where memory is scarce.
متن کاملذخیره در منابع من
با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید
عنوان ژورنال:
دوره شماره
صفحات -
تاریخ انتشار 2003