Synopsys, Inc. · NASDAQ:SNPS
Synopsys — SNUG Saber 2002
15 slides from Synopsys's conference presentation, 2002
Slide 1: Reducing electrical system design effort with automated design analysis
Neil YourTaylor, Name (and Chris company Price logo if desired — change this on slide master) Reducing electrical system design effort with automated design analysis Neil Taylor • FirstEarth Ltd Neil.Taylor@firstearth.co.uk Prof. Chris Price • University of Wales, Aberystwyth (UWA), UK cjp@aber.ac.u
Slide 2: Challenges for electrical systems design
Neil Taylor, 2 Challenges for electrical Chris Price systems design • Competing design pressures – Increasing design complexity – Commercial pressures to reduce: Time to market Development costs – Maintain and improve product quality, safety and reliability • Standard design analysis techniques, inc
Slide 3: Overview of the presentation
Neil Taylor, 3 Chris Price Overview of the presentation • Moving from manual to automated analysis • Automated analysis process • Information required • Summary of experiences • From Design to Diagnosis • Conclusions SNUG Saber 2002 3
Slide 4: Moving from manual to automated analysis
Neil Taylor, 4 Moving from manual to Chris Price automated analysis • Qualitative analysis – Early identification of analysis – Complete and consistent analysis – Unable to resolve simulations where numerical values are required e.g. will a short circuit blow the fuse? • Numerical analysis – Used la
Slide 5: Automated Design Analysis - Process Overview
Neil Taylor, 5 Automated Design Analysis - Chris Price Process Overview 1. Draw System Design 2. Define inputs and outputs 3. Simulate behaviour 5. Interactive Analysis 4. Report Generation SNUG Saber 2002 5
Slide 6: Key information
Neil Taylor, 6 Chris Price Key information • Inputs – Schematic, e.g. SaberSketch, SaberHarness – MAST Models – QSchemes - defining key bands to interpret the numerical results – Component Mappings - wrapper for the MAST model – Function models - define key system outputs • Outputs – Simulations wit
Slide 7: Interpreting Saber results with Quantization Schemes
Neil Taylor, 7 Interpreting Saber results with Chris Price Quantization Schemes • Need to recognize significant outputs • Map number ranges onto key labels SNUG Saber 2002 7
Slide 8: Component Mappings
Neil Taylor, 8 Chris Price Component Mappings • Define wrapper information for MAST models • Used for components that: – Contain inputs – Have outputs that you want to measure – Have failure modes that you want to investigate SNUG Saber 2002 8
Slide 9: Function modeling
Neil Taylor, 9 Chris Price Function modeling • Require a method to identify key outputs from a system • Define function models • Link models to schematics SNUG Saber 2002 9
Slide 10: Simulations
Neil Taylor, 10 Chris Price Simulations • Activate the key inputs to the system – Modifies the parameters to the Saber MAST models • View results - function’s achieved and back colouring SNUG Saber 2002 10
Slide 11: Failure Mode Effects Analysis (FMEA)
Neil Taylor, 11 Failure Mode Effects Analysis Chris Price (FMEA) • Start FMEA – Define scenario – Select failures • Result is textual report • Investigate each row • Re-simulate a particular row to understand issues described • Incremental FMEA SNUG Saber 2002 11
Slide 12: Sneak Circuit Analysis
Neil Taylor, 12 Chris Price Sneak Circuit Analysis • Using same underlying analysis, perform Sneak Circuit Analysis • Run with normal operation, or with failures applied SNUG Saber 2002 12
Slide 13: Benefits of automated analysis
Neil Taylor, 13 Chris Price Benefits of automated analysis • Early identification of design concerns • Comprehensive, consistent analysis and reporting • Ability to test many different design alternatives, without cost • Test virtual destruction • High confidence in electrical systems • Incremental
Slide 14: From Design to Diagnosis
Neil Taylor, 14 Chris Price From Design to Diagnosis • Re-use modelling effort for the whole lifecycle • Virtual Breadboarding – Prepare and test and debug scripts – Perform this earlier than physical breadboarding • Diagnosis – FMEA results form basis of diagnostic information FMEA is cause - effec
Slide 15: Conclusions
Neil Taylor, 15 Chris Price Conclusions • Manual analysis isn’t keeping pace with modern designs • Reduced effort through advanced simulation • Reuse existing MAST models • Additional information enables system FMEA and SCA • Techniques developed with engineer input • Opportunities for re-using mode