Agenda TRIZ Development at Intel Corporation Innovation at Intel TRIZ history at Intel
TRIZ Role and Proliferation Japan TRIZ Symposium 2008 TRIZ as an Intel methodology
Amir Roggel Principal Engineer Intel Israel
(C) The Author & Japan TRIZ Society Slide 1 Japan TRIZ Symposium 2008 September 10, 2008 (C) The Author & Japan TRIZ Society Slide 2 Japan TRIZ Symposium 2008 September 10, 2008
The 20th Century Radical Innovation ~1900 ~2000 y Goes beyond competitive positioning y May lead to major paradigm shift y Usually an individual achievement; champion driven y Risky with high failure rate; rare in mature companies
Courtesy Gene Meieran, Senior Intel Fellow Courtesy Gene Meieran, Senior Intel Fellow (C) The Author & Japan TRIZ Society Slide 3 Japan TRIZ Symposium 2008 September 10, 2008 (C) The Author & Japan TRIZ Society Slide 4 Japan TRIZ Symposium 2008 September 10, 2008 Radical Innovation Incremental Innovation 1920 y Address issues in non-traditional ways y Necessary to retain competitive position but does not 1/1,000,000 the cost 1/10,000,000 the size threaten status quo 1,000,000 X the reliability y Responsive to problems, opportunities or trends y Team driven; high expectation of success 1960 y Lots of recognition and reward for success 2000
Courtesy Gene Meieran, Senior Intel Fellow Courtesy Gene Meieran, Senior Intel Fellow (C) The Author & Japan TRIZ Society Slide 5 Japan TRIZ Symposium 2008 September 10, 2008 (C) The Author & Japan TRIZ Society Slide 6 Japan TRIZ Symposium 2008 September 10, 2008
The Future of Moore’s Law Incremental Innovation: Moore’s Law Communications No Relaxing! Design Info tech Architecture Networks Systems Automation Software Electronics Optics Mech. Eng. Photonics Design Operations Chem E. Solid state physics .09µ 0.065 106 Chemistry Physics Core2Duo® Pentium®4 Processor Core2Duo 2007 65nm 105 Materials science Pentium® II Processor ® 104 Pentium III Processor Pentium® Processor Pentium® Pro Processor 103 i486™ Processor Networking 102 i386™ Processor Manufacturing 80286 Design 10 8086 Architecture Processing 1 ’75 ’80 ’85 ’90 ’95 ’00 ’05 ’10 ’15 The Number of Transistors Per Chip The Number of Transistors Per Chip Source: Intel WillWill Double Double Every Every 18 18 Months. Months. Courtesy Gene Meieran, Senior Intel Fellow Courtesy Gene Meieran, Senior Intel Fellow (C) The Author & Japan TRIZ Society Slide 7 Japan TRIZ Symposium 2008 September 10, 2008 (C) The Author & Japan TRIZ Society Slide 8 Japan TRIZ Symposium 2008 September 10, 2008 CPU – the Brain of Evolving Computing Platforms 20th Century Top Science and
Intuitive and Aware Technology Achievements * Computers
1. Electrification 11. Interstate highways Mobile 2. Automobile 12. Space flight 3. Airplane 13. Internet Networked Sensors 4. Water supply & distribution 14. Imaging +Inference 5. Electronics 15. Household appliances 6. Radio and television 16. Health technologies Mainframes Wireless 7. Agricultural mechanization 17. Petrochemical technology 8. Computers 18. Laser and fiber optics Will sensors and inference PC’s trigger a “Cambrian” explosion? 9. Telephone 19. Nuclear technologies 10. Air conditioning/refrigeration 20. High-performance materials * National Academy of Engineering; Providing ubiquitous power was ENIAC “A Century of Innovation” a key to success in the 20th century
Courtesy Andrew Chien – VP and Director, Intel Research Courtesy Gene Meieran, Senior Intel Fellow (C) The Author & Japan TRIZ Society Slide 9 Japan TRIZ Symposium 2008 September 10, 2008 (C) The Author & Japan TRIZ Society Slide 10 Japan TRIZ Symposium 2008 September 10, 2008
The 20th Century All manufactured in 21st Century Top Science and high volume at Technology Achievements * televisionaffordable prices radio Cell phone Energy conservation Traffic/population logistics computers Electro Opto Resource protection Natural disaster control Food & water and supply AI, interfaces and robotics nuclear movies appliances laser Waste management Electronic environment Mecho Medicine and prolonging life Globalization electrification imaging Security & counter-terrorism Space exploration Education and learning Preservation of species Sciences: airplane New technology Entertainment Chemistry Depended on Genetics and cloning “Virtualization” and VR Physics energy Knowledge sharing Preservation of history Materials car mechanization distribution Global communication *Gene Meieran prediction; IFF votes
Courtesy Gene Meieran, Senior Intel Fellow Courtesy Gene Meieran, Senior Intel Fellow (C) The Author & Japan TRIZ Society Slide 11 Japan TRIZ Symposium 2008 September 10, 2008 (C) The Author & Japan TRIZ Society Slide 12 Japan TRIZ Symposium 2008 September 10, 2008 The 21st Century TRIZ at Intel All to be manufactured in high volume at y History affordable prices BIO y Proliferation Waste BIO mgmtBiology Medical INFO y Problem solving and Medicine Information y Applications: Present and Future Energy Food prod Sensors & dist Communications NANO AI Sciences: Will depend on WeatherNano -scale Physics Knowledge Biology TechnologySecurity distribution Information “Chips” Internet
(C) The Author & Japan TRIZ Society Slide 13 Japan TRIZ Symposium 2008 September 10, 2008 (C) The Author & Japan TRIZ Society Slide 14 Japan TRIZ Symposium 2008 September 10, 2008
TRIZ at Intel
y 1996-2001 Exploration stage Intel’s Manufacturing Dalian - 1996, Santa Clara Technology Development - Began TRIZ Fab/Sort & Assembly/Test Sites Fab software pilot/training. Two very successful projects – “Sputnik” and “Bubbles” - 1998 Introduced to Assembly Technology Development and Flash Business 2008 1970 Washington Systems Ireland y 2002-2004 Early deployment and seeding in Mfg. Fab - 2002 First TRIZ class in Assembly/Test Mfg. – Cavite, Philippines Chengdu*Chengdu Shanghai - 2003 First class in Fab/Sort Mfg. – Kiryat Gat, Israel Oregon Israel Mass. Assembly/Test - 2004 Classes in more sites (Fab/Sort and Assembly/Test) Fab/Dvlpmt. Mass. Fab Assembly/Test Fab ● y 2005-2006 Adoption – Manufacturing world-wide California Philippines - 2005 First classes to Level-2 and Level-3 Fab/Dvlpmt. Assembly/Test 2006 All Level-1, Level-2 classes delivered internally Costa Rica - Arizona VietNam New Mexico Assembly/Test Fab Fab Assembly/Test y 2007-> into the future A/T Dvlpmt. Malaysia - Manufacturing expansion and beyond Assembly/Test
(C) The Author & Japan TRIZ Society Slide 15 Japan TRIZ Symposium 2008 September 10, 2008 (C) The Author & Japan TRIZ Society Slide 16 Japan TRIZ Symposium 2008 September 10, 2008 Intel TRIZ Conference Mark Barkan – MA-TRIZ President Chandler, Arizona - Dec 2007 y Intel wide conference held in 2007. y Conference Theme: “Innovating the Future” y Over 150 attendees from around the world (40 non-US). y 6 papers, 28 posters showcasing work across multi-disciplines y Invited talks by Intel Fellows, industry experts, TRIZ masters
(C) The Author & Japan TRIZ Society Slide 17 Japan TRIZ Symposium 2008 September 10, 2008 (C) The Author & Japan TRIZ Society Slide 18 Japan TRIZ Symposium 2008 September 10, 2008
Problem Solving Framework at Intel Why is it important for Intel to solve manufacturing problems quickly?
Each new technology has a shorter life cycle and requires a faster ramp y Why problem solving is so rate than previous technologies. Problems incurred are much more costly per day of problem solving than they have ever been. critical in our current business model
y Our current problem solving 90nm culture 130nm
y Our manufacturing objectives 180nm
y Types of problems we solve 300mm WSPW 250nm y How TRIZ fits in y Future vision of TRIZ at Intel 350nm
65nm
(C) The Author & Japan TRIZ Society Slide 19 Japan TRIZ Symposium 2008 September 10, 2008 (C) The Author & Japan TRIZ Society Slide 20 Japan TRIZ Symposium 2008 September 10, 2008 Intel’s Manufacturing Mission Intel’s Problem Solving Culture and Methodology To be the best in the world at: Problem or Undesired event occurrence Ramping new products into high volume Yields Prevent Symptoms Fix
Tool productivity Most Probable Cause(s) Cost and agility Root Cause
In order to achieve our goals, problems must be quickly Problems are solved at different levels, with both contained, and with root cause understanding. time-to-solution and quality of solution in mind.
(C) The Author & Japan TRIZ Society Slide 21 Japan TRIZ Symposium 2008 September 10, 2008 (C) The Author & Japan TRIZ Society Slide 22 Japan TRIZ Symposium 2008 September 10, 2008
Managing the Problem Solving Process Types of Problems We Solve
Framework to manage the y Corrective: Problems where a problem solving process Step 1: standard previously achieved is Define Step 2: not being met Ex: excursion. the Problem Step 7: Document Determine Current Nexte Steps Situation y Improvement: The current system or process performance, as it was designed, is expected to be improved Ex: Yield improvement. Step 6: Communicate! Step 3: Standardize Create & Solutions Validate Model y Preventive: Problems are where the goal is to add robustness and Step 5: Step 4: prevent systems or processes from Implement Develop failure or falling below baseline. Solutions Solutions Ex: Integrated process window centering
(C) The Author & Japan TRIZ Society Slide 23 Japan TRIZ Symposium 2008 September 10, 2008 (C) The Author & Japan TRIZ Society Slide 24 Japan TRIZ Symposium 2008 September 10, 2008 Analytic Tools for Problem Solving How the Problem Solving Model Fits Together y This is where TRIZ comes in… y Ramping new products into HVM Management tools Analytical tools Corrective Problems Failure Anticipation • Functional Analysis of Product or process. Analysis Problem Definition • TRIZ helps to define the real problem to be worked on vs. symptoms. Problem Definition Improvement Problems • Ideality of Engineering Systems & focusing on Main Useful Function. Step 1: • “Zone of Conflict” Analysis of the Useful / Harmful Operational Define Step 2: Functional Modeling CurrentCurrent SituationSituation Zones to understand the conflicting requirements Step 7: the Problem Document • Resource Analysis during Operational Time Determine Current Nexte Steps Situation CEC ModelModel Building/Building/ • Cause and Effect Chain helps to see other potential causes of RootRoot CauseCause problems AnalysisAnalysis • Scientific Effects db better understanding of other potential causes Step 6: Communica Step 3: Standardize te! Create & SolutionSolution • Scientific Effects db identifying effects that can solve problems Solutions Validate Model • 40 Principles – identify potential solutions DefinitionDefinition • Predictions (Technology Trends) – identifies evolutionary potential of interaction between components Step 5: Step 4: Implement Develop Solutions Solutions
(C) The Author & Japan TRIZ Society Slide 25 Japan TRIZ Symposium 2008 September 10, 2008 (C) The Author & Japan TRIZ Society Slide 26 Japan TRIZ Symposium 2008 September 10, 2008
How the Problem Solving Model Fits How the problem solving model fits Together together y Yields y Tool productivity
Corrective Problems Corrective Problems Preventive Problems Preventive Problems
76 Inventive Functional Modeling Improvement Problems Step 1: Standards Improvement Problems Step 1: Define Define Step 2: Step 2: the Problem the Problem Trimming Step 7: Document ARIZ Step 7: Document Determine Current Determine Current Nexte Steps Situation Nexte Steps Situation CEC
Step 6: Communicate! Step 3: Step 6: Communicate! Step 3: Standardize Create & Standardize Create & Solutions Validate Model Solutions Validate Model
Step 5: Step 4: Step 5: Step 4: Implement Develop Implement Develop Solutions Solutions Solutions Solutions
(C) The Author & Japan TRIZ Society Slide 27 Japan TRIZ Symposium 2008 September 10, 2008 (C) The Author & Japan TRIZ Society Slide 28 Japan TRIZ Symposium 2008 September 10, 2008 How the Problem Solving Model Fits Accomplishment and Challenges Together y Cost and agility y TRIZ has generated significant benefit for Intel as measured by time to solutions and cost saving y Challenges in deployment of TRIZ in Intel - Strengthening existing uses Improvement Problems - Propagate to new areas of use y Tips for other companies who want to deploy TRIZ Step 1: Process Analysis Define Step 2: - TRIZ is not “Magic Wand”: it requires hard work and the Problem Step 7: Document investment, and deliver great results to those who do it Determine Current Nexte Steps Situation - The 4 conditions required to make TRIZ successful in a company…
Step 6: Communicate! Step 3: Standardize Create & Solutions Validate Model
Step 5: Step 4: Implement Develop Solutions Solutions
(C) The Author & Japan TRIZ Society Slide 29 Japan TRIZ Symposium 2008 September 10, 2008 (C) The Author & Japan TRIZ Society Slide 30 Japan TRIZ Symposium 2008 September 10, 2008
Conclusions Acknowledgements
y TRIZ is a key systematic innovation platform for Intel into 21st century y Our manufacturing challenge: Improve productivity and efficiency without compromise y Application of TRIZ in R&D and management Mike Rocke are evolving: to improve and develop Kevin Brune Dave Troness Richard Platt systems and processes throughout Intel y TRIZ offers both tactical and strategic Our Visionaries and Champions capability for companies. It is only the beginning…
(C) The Author & Japan TRIZ Society Slide 31 Japan TRIZ Symposium 2008 September 10, 2008 (C) The Author & Japan TRIZ Society Slide 32 Japan TRIZ Symposium 2008 September 10, 2008 David Austin TS Yeoh Principal Engineer Principal Engineer Intel Arizona Intel Penang
Co Authors/leaders
Our Teachers Our Sponsor
Sergei Ikovenko Alex Lyubomyskiy Gene Meieran (C) The Author & Japan TRIZ Society Slide 33 Japan TRIZ Symposium 2008 September 10, 2008