Automatic boosting of cross-product coverage using Bayesian networks

Baras, Dorit; Fine, Shai; Fournier, Laurent Sébastien; Geiger, Dan; Ziv, Avi

doi:10.1007/s10009-010-0160-z

Cited by 9 publications

(6 citation statements)

References 10 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In addition, the methods need to become accessible to a broader audience, one that does not necessarily need to be literate in machine learning. So far, the approaches that are closer to achieving the aforementioned target are those of Baras et al [2008] and Bernardi et al [2008].…”

Section: Discussionmentioning

confidence: 97%

“…Then, given an uncovered functional coverage point (or group of them), queries would be formed to elicit from the BN model which directives would be the most probable to use in covering those points. The DUVs used were both microprocessor designs [Fine et al 2006] as well as smaller internal peripheral modules [Baras et al 2008;Braun et al 2004].…”

Section: Probabilistic Methods (Bayesian Network Markov Models) a mentioning

confidence: 99%

“…The latest effort of the research team [Baras et al 2008] focuses on automating the DAG construction phase. As the approach in Fine and Ziv [2003a] required considerable engineering effort and expertise to construct the main BN, the newly proposed system uses several generic structure learning algorithms, which are later enhanced by two proposed heuristics to significantly reduce the need for manual involvement.…”

Section: :11mentioning

confidence: 99%

See 2 more Smart Citations

Coverage-Directed Test Generation Automated by Machine Learning -- A Review

Ioannides

Eder

2012

ACM Trans. Des. Autom. Electron. Syst.

View full text Add to dashboard Cite

The increasing complexity and size of digital designs, in conjunction with the lack of a potent verification methodology that can effectively cope with this trend, continue to inspire engineers and academics in seeking ways to further automate design verification. In an effort to increase performance and to decrease engineering effort, research has turned to artificial intelligence (AI) techniques for effective solutions. The generation of tests for simulation-based verification can be guided by machine-learning techniques. In fact, recent advances demonstrate that embedding machine-learning (ML) techniques into a coverage-directed test generation (CDG) framework can effectively automate the test generation process, making it more effective and less error-prone. This article reviews some of the most promising approaches in this field, aiming to evaluate the approaches and to further stimulate more directed research in this area.

show abstract

Section: Discussionmentioning

confidence: 97%

Section: Probabilistic Methods (Bayesian Network Markov Models) a mentioning

confidence: 99%

Section: :11mentioning

confidence: 99%

See 1 more Smart Citation

Coverage-Directed Test Generation Automated by Machine Learning -- A Review

Ioannides

Eder

2012

ACM Trans. Des. Autom. Electron. Syst.

View full text Add to dashboard Cite

show abstract

“…Often Coverage Driven Verification (CDV) is used to determine the targeted tests that need to be written. [1,2] Random tests, on the contrary, often have the ISA instruction set completely open and may mix random biases (such as dependencies, resource contention, exceptions, etc.) into the mix to provide a broad test with less depth.…”

Section: Testsmentioning

confidence: 99%

Coverage Learned Targeted Validation for Incremental HW Changes

Farkash¹,

Hickerson²,

Behm³

2014

Proceedings of the 51st Annual Design Automation Conference

View full text Add to dashboard Cite

This paper addresses the challenges of minimizing the time and resources required to validate the changes between two Hardware (HW) model iterations of the same design. It introduces CLTV (Coverage Learned Targeted Validation), an automatic framework which learns during the verification process of the HW and uses the learned information to target the areas of the design that are affected by the incremental HW model iterations.Our paper defines new concepts, presents our implementation of the supporting algorithms, and shows actual results on an IBM POWER8 processor with outstanding results.

show abstract

“…probability distributions used in the randomization process have been left out. To address this, a few studies [Bar08], [Fin09] adopted a probabilistic approach but they failed to mention actual implementation in production cycle and scalability issue. The majority of the previous research on hardware verification with the simulation-based testing approach has focused on supervised learning [Mam16], [Bar08], [Wag07] and evolutionary algorithms [Ber13], [Cru13].…”

Section: Previous Machine-learning Based Approachmentioning

confidence: 99%

Case study: Real-world machine learning application for hardware failure detection

Shin¹

2019

Proceedings of the Python in Science Conference

View full text Add to dashboard Cite

When designing microprocessors, engineers must verify whether the proposed design, defined in hardware description language, does what is intended. During this verification process, engineers run simulation tests and can fix bugs if the tests have failed. Due to the complexity of the design, the baseline approach is to provide random stimuli to verify random parts of the design. However, this method is time-consuming and redundant especially when the design becomes mature and thus failure rate is low. To increase efficiency and detect failures faster, it is possible to train machine learning models by using previously run tests, and assess the likelihood of failure of new test candidates before running them. This way, instead of running random tests agnostically, engineers use the model prediction on a new set of test candidates and run a subset of them (i.e., "filtering" the tests) that are more likely to fail. Due to the severe imbalance (1% failure rate), I trained an ensemble of supervised (classification) and unsupervised (outlier detection) models and used the union of the prediction from both models to catch more failures. The tool has been deployed in an internal high performance computing (HPC) cluster early this year, as a complementary workflow which does not interfere with the existing workflow. After the deployment, I found performance instability in post-deployment performance and ran various experiments to address the issue, such as by identifying the effect of the randomness in the test generation process. In addition to introducing the relatively new data-driven approach in hardware design verification, this study also discusses the details of post-deployment evaluation such as retraining, and working around real-world constraints, which are sometimes not discussed in machine learning and data science research.

show abstract

Automatic boosting of cross-product coverage using Bayesian networks

Cited by 9 publications

References 10 publications

Coverage-Directed Test Generation Automated by Machine Learning -- A Review

Coverage-Directed Test Generation Automated by Machine Learning -- A Review

Coverage Learned Targeted Validation for Incremental HW Changes

Case study: Real-world machine learning application for hardware failure detection

Contact Info

Product

Resources

About