EDA
System
IP
SolvNetPlus
Training & Education
Overview
As modern SoC designs increase in complexity, debugging X-propagation issues has become a critical bottleneck in verification workflows. The GPU IP development and validation team at Intel encountered challenges in isolating root causes of X-related assertion failures, where traditional methods resulted in long debug cycles and inefficient engineering effort. By adopting Synopsys Verdi® XRCA (Root Cause Analysis) technology, Intel transformed its debug process—accelerating turnaround time, improving productivity, and enabling faster verification closure.
"XRCA has fundamentally changed how we approach X-propagation debug. What once took days of iterative analysis can now be resolved in just a few hours. The automation and root-cause correlation capabilities have significantly improved our productivity and helped us accelerate verification closure."
Debabrata Chatterjee
|Senior Principal Engineer at Intel
Challenges
X-propagation events—caused by undriven signals, dangles, or low-power effects—introduce unknown (X’) values that propagate through the design and trigger cascades of assertion failures. These failures generate significant debug noise, making it difficult to pinpoint root causes.
Intel’s teams relied on a sequential debug approach: analyzing the first failure, applying a fix, rerunning simulations, and repeating the process for subsequent failures. This iterative workflow increased debug turnaround time, consumed valuable engineering resources, and posed risks to project schedules.
Solution
Intel deployed Synopsys Verdi XRCA technology to automate and scale X-propagation debug. XRCA delivers powerful capabilities that fundamentally improve debug efficiency:
- Automated root-cause analysis that traces assertion failures back to their source using logs and waveform data
- Parallel debugging to analyze and resolve multiple failures simultaneously
- Machine learning–driven clustering to correlate failures and identify common root causes
- Broad applicability across RTL, gate-level, low-power, and X-pessimism scenarios
With configurable inputs, XRCA generates detailed trace reports, enabling engineers to visualize propagation paths and quickly resolve issues.
Results
By integrating XRCA into its verification flow, Intel achieved dramatic improvements in debug throughput and turnaround time which was presented at SNUG Silicon Valley 2026:
- Reduced single-assertion debug time from ~6 hours to ~0.5 hours
- Reduced multi-assertion debug time from several days to ~3 hours
- Increased overall debug productivity and reduced engineering effort
- Established XRCA as the standard methodology for X-propagation debug
Synopsys Verdi XRCA enables a step-function improvement in X-propagation debug by combining automation, parallel analysis, and machine learning-driven insights. Intel’s adoption of Verdi XRCA demonstrates how advanced debug technologies can simplify complex verification challenges, reduce time-to-resolution, and improve overall design quality.
