Summary of Improving Simulation Regression Efficiency Using a Machine Learning-based Method in Design Verification, by Deepak Narayan Gadde et al.
Improving Simulation Regression Efficiency using a Machine Learning-based Method in Design Verification
by Deepak Narayan Gadde, Sebastian Simon, Djones Lettnin, Thomas Ziller
First submitted to arxiv on: 24 May 2024
Categories
- Main: Machine Learning (cs.LG)
- Secondary: Hardware Architecture (cs.AR)
GrooveSquid.com Paper Summaries
GrooveSquid.com’s goal is to make artificial intelligence research accessible by summarizing AI papers in simpler terms. Each summary below covers the same AI paper, written at different levels of difficulty. The medium difficulty and low difficulty versions are original summaries written by GrooveSquid.com, while the high difficulty version is the paper’s original abstract. Feel free to learn from the version that suits you best!
| Summary difficulty | Written by | Summary |
|---|---|---|
| High | Paper authors | High Difficulty Summary Read the original abstract here |
| Medium | GrooveSquid.com (original content) | Medium Difficulty Summary The paper presents various methods for improving verification throughput in System-on-Chip (SoC) designs. The primary challenge is the increasing complexity and size of SoC designs, making it difficult to verify them efficiently using traditional methods. Two approaches are discussed: ranking and Xcelium ML, a machine learning-based technology introduced by Cadence. Ranking selects seeds that achieve high coverage in previous simulations, while Xcelium ML generates optimized patterns based on correlations between randomization points and achieved coverage. The paper compares the two methods using quantified results from three industry projects, demonstrating comparable compression and speedup factors around 3 for both approaches. However, Xcelium ML occasionally produces a significant regain of more than 100% in coverage when simulating new random scenarios. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary The paper talks about making sure computer chips work correctly before they’re made. It’s getting harder to do this because the chips are getting really complex and big. The authors suggest two ways to make it faster: ranking and a new machine learning technique called Xcelium ML. Ranking picks the most important things to test, while Xcelium ML looks for patterns in how well tests work and uses that information to make better tests. The paper shows examples of three real-life projects where both methods worked similarly well, but Xcelium ML sometimes found even more problems. |
Keywords
» Artificial intelligence » Machine learning
