|
Statistical variation causes product failure. Analog/mixed-signal, custom digital and memory IC designers are squeezed by aggressive product cycles, low supply voltages, low power and increasing variation. The unspoken truth is that most products tape-out with serious variation issues. How do you find the source of failure? What should you change in the design to fix it? Products suffer from low margin and delays in time-to-market. Process technologies and circuit design techniques have made incredible progress, but tools for coping with variation have not kept pace. Best- and worst-case corners only treat global variation. What about local variation/mismatch, which is often dominant? Monte Carlo simulation is time-consuming and inefficient; surely there is a better solution. Designing for statistical variation needs to be faster and produce better results. SolidoSTAT is a 2nd generation statistical variation design tool that overcomes the limitations of 1st generation tools like Monte Carlo. With SolidoSTAT, you can quickly analyze failures due to statistical variation, identify design weaknesses, and fix the design, making it robust to statistical variation. Analyze......failures due to statistical variation.Monte Carlo is too slow. SolidoSTAT finds failures far faster than Monte Carlo, giving you more time to identify and fix problems precisely. Using 2nd generation sampling techniques such as Latin Hypercube Sampling, Tail Sampling™, Run-Time Feedback and True Corners™ discovery, SolidoSTAT efficiently analyzes your design under the most important process conditions. Efficiency gains for typical designs range between 10x and 1,000x. This means less time waiting for results, and better use of simulator licenses and compute resources. Identify......weaknesses in the design.Monte Carlo doesn’t identify the cause of failures. SolidoSTAT goes beyond the histograms and polynomial response surface models used in 1st generation tools, pinpointing key weaknesses that cause design failure and providing guidance for fixing them. Unlike other tools, confidence intervals are provided with results, instead of presenting best guesses as fact. SolidoSTAT identifies problematic sources of variation, critical devices, and limiting environmental conditions, so that you can focus on fixing problems rather than finding them. Fix......the design, making it robust to statistical variation.With SolidoSTAT, you can fix design failures in minutes instead of days. SolidoSTAT eliminates the need for tedious iteration between design changes and Monte Carlo analysis. Instead, SolidoSTAT lets you fix design weaknesses by rapidly iterating with a small set of True Corners. Critical circuit blocks can be sized for the ideal combination of robustness, performance, and area using detailed block sizing capabilities instead of time-consuming hand analysis. Robust sizing options for the circuit can also be generated automatically. The result is a high-quality, cost-effective, variation-robust design. |
Statistical variation causes failures. 
2nd generation statistical variation design is faster and more efficient than Monte Carlo. |
 |
Analyzefailures due to statistical variation |
Identifyweaknesses in the design |
Fixthe design, making it robust to statistical variation |
Fast Statistical Sampling:
|
Statistical Characterization identifies causes of design failure:
|
Block Sizing: quickly size critical circuit blocks for the ideal combination of robustness, performance and area Rapid Sizing: rapidly size your design with True Corners™, avoiding the need for tedious iteration between design changes and Monte Carlo analysis Circuit Enhancement: automatically resize your design for statistical variation robustness |
|
 |
|
|
|
Additional Features
|
Customer Case Studies
|
