PowerQuest: Trace Driven Data Mining for Power Optimization

Babighian, Pietro; Kamhi, Gila; Vardi, Moshe Y.

doi:10.1109/date.2007.364437

Cited by 12 publications

(2 citation statements)

References 13 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In Figure 4, results gained for a subset of ITC99 benchmarks are presented and compared to reduction (r1) published in [2] and [3], which are denoted as method A and B in the figure. It is evident reduction got by our method is better than reduction got by the mentioned methods …”

Section: Cmentioning

confidence: 99%

The Use of Genetic Algorithm to Derive Correlation Between Test Vector and Scan Register Sequences and Reduce Power Consumption

Kotásek

karvada

Strnadel

2010

2010 13th Euromicro Conference on Digital System Design: Architectures, Methods and Tools

View full text Add to dashboard Cite

In most of existing approaches, the reorganization of test vector sequence and reordering scan chains registers to reduce power consumption are solved separately, they are seen as independent procedures. In the paper it is shown that a correlation between these two processes and strong reasons to combine them into one procedure run concurrently exist. Based on this idea, it is demonstrated that search spaces of both procedures can be combined together into a single search space in order to achieve better results during the optimization process. The optimization over the united search space was tested on ISCAS85, ISCAS89 and ITC99 benchmark circuits implemented by means of CMOS primitives from AMI technological libraries. Results presented in the paper show that lower power consumption can be achieved if the correlation is reflected, i.e., if the search space is united rather than divided into separate spaces. At the end of the paper, results achieved by genetic algorithm based optimization are presented, discussed and compared with results of existing methods. (Abstract)

show abstract

Section: Cmentioning

confidence: 99%

The Use of Genetic Algorithm to Derive Correlation Between Test Vector and Scan Register Sequences and Reduce Power Consumption

Kotásek

karvada

Strnadel

2010

2010 13th Euromicro Conference on Digital System Design: Architectures, Methods and Tools

View full text Add to dashboard Cite

show abstract

“…Since this is equivalent to classification of the given objects in accordance to the set of features, this process (class label extraction) identifies machine learning based classification techniques. There has been several attempts to apply trace based machine learning to CAD domains in context dynamic invariant extraction [10,11].…”

Section: Related Workmentioning

confidence: 99%

Magenta

Kamhi

Novakovsky

Tiemeyer

et al. 2009

Proceedings of the 46th Annual Design Automation Conference

Self Cite

View full text Add to dashboard Cite

Adopting an ESL based design and validation methodology, we introduce a top-down approach for efficient debugging of microarchitectural specification and RTL implementation. Our solution is based on the formalism introduced by statistical transactional analysis that we call MAGENTA -Modeling AGENT for Transactional Analysis. To the best of our knowledge, MAGENTA based root-cause analysis pioneers in the efficient characterization of the micro-architectural design misbehavior via abstraction of validation output by transactions and microarchitectural events.

show abstract

Applications

2016

Boolean Circuit Rewiring

View full text Add to dashboard Cite

PowerQuest: Trace Driven Data Mining for Power Optimization

Cited by 12 publications

References 13 publications

The Use of Genetic Algorithm to Derive Correlation Between Test Vector and Scan Register Sequences and Reduce Power Consumption

The Use of Genetic Algorithm to Derive Correlation Between Test Vector and Scan Register Sequences and Reduce Power Consumption

Magenta

Applications

Contact Info

Product

Resources

About