NDLI: Role of Reduced Inputs in Flag Mining

Please wait, while we are loading the content...

Semantic Clustering Driven Approaches to Recommender Systems

An Automated Approach for Cross-Browser Inconsistency (XBI) Detection

Efficient Multi-depth Querying on Provenance of Relational Queries Using Graph Database

Attention Sensitive Web Browsing

Concept Preserving Visual Summarization of Social Image Search Results

A Novel Approach to Big Data Veracity using Crowdsourcing Techniques and Bayesian Predictors

Service Demand Modeling and Prediction with Single-user Performance Tests

Selection of Image Blocks using Genetic Algorithm and Effective Embedding with DCT for Steganography

Distributed Decision Tree

Comparative Evaluation of Various Indexing Techniques of Geospatial Vector Data for Processing in Distributed Computing Environment

Workload Aware Hybrid Partitioning

EEQuest: An Event Extraction and Query System

Role of Reduced Inputs in Flag Mining

Hierarchical Tree-based Optimized Communication for Real Time Event Driven Internet of Things

Quantitative Estimation of Cost Drivers for Intermediate COCOMO towards Traditional and Cloud Based Software Development

SkipLPA: An Efficient Label Propagation Algorithm for Community Detection in Sparse Network

Three Steps Secure Login: A systematic approach

Centrality Approach for Community Detection in Large Scale Network

Development of Indian Weighted Diabetic Risk Score (IWDRS) using Machine Learning Techniques for Type-2 Diabetes

Topical Authoritative Answerer Identification on Q&A Posts using Supervised Learning in CQA Sites

A Methodology to Detect and Track Breaking News on Twitter

Search Based Test Data Generation: A Multi Objective Approach using MOPSO Evolutionary Algorithm

Role of Reduced Inputs in Flag Mining

Content Provider	ACM Digital Library
Author	Ravindar, Archana Bansal, Rishab
Abstract	Typically compilers provide a wide choice of optimization flags that can be used to improve the application performance. The process of searching for the best flag combination for a given application is referred to Flag Mining. Brute force ways of flag mining are time consuming as it requires a number of runs with different combinations of flags. Flag mining techniques that are based on machine learning rely on a database consisting of measurements of application run-times obtained with a large number of combinations of binaries compiled with different flags. This work quantifies the impact of using reduced inputs in flag mining. Reduced inputs are much smaller inputs than real representative inputs and cause the application to run for less than 10 percent of original execution time. Some examples of reduced inputs are the train input used in SPEC benchmarks, MinneSPEC inputs. Using reduced inputs instead of full inputs would reduce time/space overhead of flag mining significantly when used in brute force or machine learning based methods. However inorder to use reduced inputs for flag mining, the behavior of the application compiled with a set of flags, when presented with reduced inputs should give similar benefits on full representative inputs. This can happen only if reduced inputs are an accurate representatives of ref inputs in the context of application performance. Our experiments show that reduced inputs correlate to full representative inputs for 5 out of 7 SPEC CPU2006 benchmarks on all 11 flag combinations considered with the GCC compiler and are found to reduce the experimentation time of flag mining by up to 82%. We also outline the necessary conditions that need to be satisfied by reduced inputs to qualify for use in flag mining.
Starting Page	67
Ending Page	76
Page Count	10
File Format	PDF
ISBN	9781450348089
DOI	10.1145/2998476.2998485
Language	English
Publisher	Association for Computing Machinery (ACM)
Publisher Date	2016-10-21
Publisher Place	New York
Access Restriction	Subscribed
Subject Keyword	Reduced inputs Representative inputs Benchmarks Flag mining Compiler optimizations Performance Spec
Content Type	Text
Resource Type	Article

Sl.	Authority	Responsibilities	Communication Details
1	Ministry of Education (GoI), Department of Higher Education	Sanctioning Authority	https://www.education.gov.in/ict-initiatives
2	Indian Institute of Technology Kharagpur	Host Institute of the Project: The host institute of the project is responsible for providing infrastructure support and hosting the project	https://www.iitkgp.ac.in
3	National Digital Library of India Office, Indian Institute of Technology Kharagpur	The administrative and infrastructural headquarters of the project	Dr. B. Sutradhar bsutra@ndl.gov.in
4	Project PI / Joint PI	Principal Investigator and Joint Principal Investigators of the project	Dr. B. Sutradhar bsutra@ndl.gov.in Prof. Saswat Chakrabarti will be added soon
5	Website/Portal (Helpdesk)	Queries regarding NDLI and its services	support@ndl.gov.in
6	Contents and Copyright Issues	Queries related to content curation and copyright issues	content@ndl.gov.in
7	National Digital Library of India Club (NDLI Club)	Queries related to NDLI Club formation, support, user awareness program, seminar/symposium, collaboration, social media, promotion, and outreach	clubsupport@ndl.gov.in
8	Digital Preservation Centre (DPC)	Assistance with digitizing and archiving copyright-free printed books	dpc@ndl.gov.in
9	IDR Setup or Support	Queries related to establishment and support of Institutional Digital Repository (IDR) and IDR workshops	idr@ndl.gov.in

Execution characteristics of SPEC CPU2000 benchmarks: Intel C++ vs. Microsoft VC++

Truth in SPEC benchmarks

The effect of compiler-flag tuning on SPEC benchmark performance

P.J.: Microarchitecture sensitive empirical models for compiler optimizations

Microarchitecture sensitive empirical models for compiler optimizations (2007)

Rating Compiler Optimizations for Automatic Performance Tuning

Cache Performance for Selected SPEC CP Benchmarks (2001)

Rapidly Selecting Good Compiler Optimizations using Performance Counters

Comparative characterization of SPEC CPU2000 and CPU2006 on Itanium® architecture

Role of Reduced Inputs in Flag Mining

Similar Documents

Execution characteristics of SPEC CPU2000 benchmarks: Intel C++ vs. Microsoft VC++

Truth in SPEC benchmarks

The effect of compiler-flag tuning on SPEC benchmark performance

P.J.: Microarchitecture sensitive empirical models for compiler optimizations

Microarchitecture sensitive empirical models for compiler optimizations (2007)

Rating Compiler Optimizations for Automatic Performance Tuning

Cache Performance for Selected SPEC CP Benchmarks (2001)

Rapidly Selecting Good Compiler Optimizations using Performance Counters

Comparative characterization of SPEC CPU2000 and CPU2006 on Itanium® architecture

Role of Reduced Inputs in Flag Mining