Pdf Solutions, Inc. · NASDAQ:PDFS
Pdf Solutions — Investor Day (November 2016)
145 slides from Pdf Solutions's investor day presentation, Nov 1, 2016
Slide 1: Analyst Day
Analyst Day November 1, 2016 John Kibarian, Ph. D. President, Chief Executive Officer, Co-Founder This presentation contains forward-looking statements regarding PDF Solutions’ future products and business prospects that involve risk and uncertainty. Actual results could differ materially from those
Slide 2: Agenda
Agenda Market Overview- John Kibarian (1:00-1:30pm) Foundational Technology- Andrzej Strojwas (2:00-2:30pm) Products and Solutions- Kimon Michaels (2:30-2:50pm) Break (2:50pm-3:00pm) Integrated Yield Ramp (IYR)- Cees Hartgring (3:00-3:20pm) Design for Inspection (DFI)- Dennis Ciplickas (
Slide 3: PDF Solutions – Mission and Vision
PDF Solutions – Mission and Vision Our Mission: Be the leading provider of innovative solutions that enable our customers to optimize process, design and fabrication of semiconductor devices for high yield, low cost and high performance Our Vision: Drive Process Control and Yield Management with
Slide 4: Mission Applies to Fabless and Foundry Eco-System
Mission Applies to Fabless and Foundry Eco-System PDF Solutions provides fabless customers information to successfully insert products into foundries and sustain them there over the product life cycle “Enabling the Virtual IDM” PDF Solutions helps foundry customers get manufacturing IC yields as hig
Slide 5: PDF Solutions – Who we are
PDF Solutions – Who we are San Jose (HQ) China San Diego Dallas Italy Pittsburgh Germany South France Korea Japan Taiwan Singapore ~425 Employees >90% in engineering ~100 PhDs Strong China Presence 10+ Years 140+ Engineers 5 / EXTERNAL USE ©(c© 2016 PDF Solutions, Inc. All rights reserved.
Slide 6: Fabless – Fabs – Foundries – OSATs – System Houses
Fabless – Fabs – Foundries – OSATs – System Houses 6 of Top-6 foundries running PDF technology 5 of Top-6 fabless companies 90 companies in 20 countries 100 Billion CV tests annually >21K process tools under PDF process control PDF Headquarters PDF Support/Development Centers Representative industry
Slide 7: Level setting: Short Term Trends
Level setting: Short Term Trends 2017 leading edge • • 14nm/16nm will be primary leading edge node First products will ramp in 10nm, likely a minor node. 10nm 28nm • • • More derivatives available: FD, HPC+, eMemory, eRF, etc China 28nm volumes not significant until late 2017 28nm production pea
Slide 8: Industry Trends: Consolidation in the West, Expansion in China
Industry Trends: Consolidation in the West, Expansion in China Declining growth in west in part has led to consolidation Seller Buyer $B Year Novellus Lam $3B 2011 National TI $7B 2011 LSI Avago $7B 2013 Freescale NXP $12B 2015 BRCM Avago $37B 2015 Altera Intel $17B 2015 PMC Skyworks $2B 2015 HMI
Slide 9: Moore's Law Changing
Moore’s Law Changing Historically, Moore’s law created a steady cycle of profitability that was apportioned between fabless, foundry, and equipment suppliers From the 1960’s Moore’s Law was achieved from geometry scaling Geometry scaling is becoming more challenging and slowing Consumer dema
Slide 10: Computing Has Been Improving Exponentially for Decades
Computing Has Been Improving Exponentially for Decades Source: Jonathan G. Koomey, Ph.D., Consulting Professor, Stanford University 2010 Exponential benefits achieved with a variety of technologies: • vacuum tubes • discrete transistors • integrated bipolar circuits • CMOS monolithic micro proce
Slide 11: Implications of Semiconductor Industry Changes
Implications of Semiconductor Industry Changes The Big are getting bigger Chip companies are integrating more functions and becoming more system providers; system companies have become chip companies Industry growth is being driven by Asian markets and in particular China More 3-D • 3-D Memo
Slide 12: Industry Transitions have always represented big opportunities
Industry Transitions have always represented big opportunities Years 1991-95 1996-2001 2002-2007 2008-2014 Transition DARPA funds Sony/Toshiba Japanese pivot to Foundry Fabless software for IC partner for consumer interface primarily manufacturing Playstation electronics away with Exensio® from DRAM
Slide 13: 2015-20XX PDF Solutions Expands Its Role
2015-20XX PDF Solutions Expands Its Role Transition • China, System-Fabless, Electrical Scaling Opportunities • IYR in China, big-data analytics, electrical inspection PDF Technology • Exensio®, Design-for-inspection™ Market opportunity • >10 new fabs built or announced in China; comparable
Slide 14: Our Proprietary Data is Our Heartbeat
Our Proprietary Data is Our Heartbeat PDF Solutions chip designs We are able to define new offerings by unique insights gained from looking across the industry using our Focused R&D designs Large datasets Insights 14 / EXTERNAL USE ©(c© 2016 PDF Solutions, Inc. All rights reserved.
Slide 15: Technology investment is why we have unique data
Technology investment is why we have data unique data pdFasTest® eProbe™ E-Test tool E-beam tool On-wafer DirectProbe™ 0 1 2 3 4 5 6 8 9 1011121314 CV® tester In-chip CV® 16 17 1819202122 24 25 2627282930 32 33 3435363738 40 41 4243444546 48 49 5051525354 31 31 31 (array, CV® 56 57 5859606162 turn p
Slide 16: PDF Solutions – Where are we going
PDF Solutions – Where are we going Product Industry Dynamic PDF Solutions Response Business impact Integrated Yield Ramp Next -> 10/7nm Expand ramps in China 7nm contract Q1/16 for foundry/Fabless New 28nm/16nm ramps fabless/fabs leveraging our Numerous Chinese in China large local engineering team
Slide 17: Summary
Summary Opportunity is expanding • Fabless and Fab build out in China • Need for advanced production control for fabless and fabs • Electrical inspection PDF Solutions investments position it well to capture the opportunities at hand • Investments made over the last decade • Core technology and
Slide 18: Thank You
THANK YOU
Slide 19: PDF Solutions' Technology Foundations
PDF SOLUTIONS’ TECHNOLOGY FOUNDATIONS Andrzej J. Strojwas, Ph. D. Chief Technologist This presentation contains forward-looking statements regarding PDF Solutions’ future products and business prospects that involve risk and uncertainty. Actual results could differ materially from those discussed. Y
Slide 20: The Design to Manufacturing Relationship
The Design to Manufacturing Relationship In the good old days… Process Development Kit Spice Models Manufacturing complexity hidden 20 / EXTERNAL USE ©(c© 2016 PDF Solutions, Inc. All rights reserved.
Slide 21: …is no longer a simple hand-off
…is no longer a simple hand-off It is extremely challenging to predict and model all impacts of design choices a priori. Detailed characterization is necessary and even then, too many choices (models) Now… Interdependent rules, RET, printability, small process windows • Design Rules, OPC, Fill, Ma
Slide 22: Technology Evolution: Process Complexity Increase
Technology Evolution: Process Complexity Increase 14nm SOC Technology: Second generation 3- dimensional FinFET device architecture 13 Cu metal layers More than 90 masks Huge increase in cycle time (up to 4 months) • Extremely long learning cycles due to multiple patterning Vanishing proces
Slide 23: Design Rule and Layout Pattern Count Explosion
Design Rule and Layout Pattern Count Explosion Tens of millions patterns •SoC Product Design Rule Count Libraries based on DR’s 100ks patterns 4000 2000 Templates Layout pattern DR’s >3,000 DRs 0 Process integration-driven DR’s 10’s of DRs 23 / EXTERNAL USE ©(c© 2016 PDF Solutions, Inc. All rights r
Slide 24: Design-Process Interface – PDF Solutions' Template
Design-Process Interface PDF Solutions’ Template™ Traditional interface between Synthesized manufacturing and design groups Designs based on design rules and SPICE Std. Cell Libraries, models is broken IP Cores DRs a prescriptive rule set: what Template™ you can’t do Technology Design Rules,
Slide 25: Yield Loss Mechanism Evolution
Yield Loss Mechanism Evolution Random Defect limited Yield decreased yield losses used to exponentially with Chip dominate: Area: Yield 0.8 Poisson 0.7 Moore Seeds 0.6 Binomial 0.5 0.4 0.3 Opens Shorts 0.2 2 Area [ cm ] 0.1 0.2 0.4 0.6 0.8 1.0 1.2 1.4 25 / EXTERNAL USE ©(c© 2016 PDF Solutions, I
Slide 26: Yield Loss Mechanism Evolution
Yield Loss Mechanism Evolution Systematic defect impact Extremely tight increased dramatically overlay/alignment specs: due to vanishing process windows Misalignment Product Yield Printability challenges due to large number of product layout patterns - 0 + Misalignment 26 / EXTERNAL USE ©(c© 201
Slide 27: Yield Loss Mechanism Evolution
Yield Loss Mechanism Evolution Random Dopant Fluctuations Line Edge Roughness -6.00 -6.50 Env I Env II Env III Layout Pattern -7.00 Dependent -7.50 -8.00 Ioff (Printability and Stress/Strain -8.50 -9.00 Parametric Effects -9.50 Effects) -10.00 -10.50 300.00 400.00 500.00 600.00 700.00 800.00 27 / EX
Slide 28: Yield Loss Mechanism Evolution
Yield Loss Mechanism Evolution Within Wafer Spatial Low yielding wafers Distributions Dominate spatial distributions: Variability and Yield Losses: Yield High Implication: Low Many points across the wafer must be measured Typical (Approximately Radial) to capture these spatial Spatial Distribu
Slide 29: Summary of Industry Challenges Driving PDF Solutions
Summary of Industry Challenges Driving PDF Solutions Process Complexity increase demands comprehensive electrical defect characterization covering all dominant yield loss mechanisms across the entire technology lifecycle • New product introduction is extremely challenging; all key design-process i
Slide 30: Test Chip Requirements at Advanced Nodes
Test Chip Requirements at Advanced Nodes Characteristics Implications Mimic Product Need wide range of layout styles Layouts 1000’s of experiments per die High Observability Ability to measure ppb failures in 3-6 wafers Within wafer variability must be captured Run subset of Judicious pruning of pro
Slide 31: PDF Solutions Characterization Vehicle (CV) Infrastructure
PDF Solutions Characterization Vehicle® (CV®) Infrastructure Electrical Fabless characterization Foundry CV ® Design IP & analytics Electrical Template™ & Fire™ measurement pdFasTest® Big data analytics Exensio® 31 / EXTERNAL USE ©(c© 2016 PDF Solutions, Inc. All rights reserved.
Slide 32: Why is a New In-line Process Control Paradigm Needed?
Why is a New In-line Process Control Paradigm Needed? Defects in advanced manufacturing Major yield issues are technologies are driven by interaction systematic, not random between extremely tight process windows and manufacturing variations Defects are becoming 3-D in nature, Must evaluate
Slide 33: What is Design-for-Inspection (DFI)?
What is “Design-for-Inspection (DFI™)?” DFI™ exploits the unused space inherent in any SoC logic design DFI™ has no area or performance penalty DFI™ Filler Cells are designed to fit seamlessly with fabless logic IP Proprietary E-beam tool enables electrically Map of available unused relevant
Slide 34: The Need for New Control Loops in Volume Production
The Need for New Control Loops in Volume Production (New Links) Fab alarms, Scanner APC FDC Inline APC Model Control Inspection Exensio® Platform Operations, Step-up plans overlay Building Sampling DFI™ Plans Post litho eProbe™ Overlay Measurement metrology pdFasTest® CV® Electrical Litho Meas. Etch
Slide 35: Yield Learning Rate
Yield Learning Rate Defectivity Reduction Rate YIELD PROCESS LIFE CYCLE VOLUME Defect Density (faults/cm2) 100% 100 3 90% 90 80% 80 2.5 70% 70 60% 60 2 MASS MANUFACTURING 50% 50 WAFER 1.5 VOLUME 40% VOLUME 40 30% 30 20% 20 1 Di 10% 10 0.5 Di+1 0% YEAR1 YEAR2 YEAR3 YEAR4 YEAR5 YEAR6 YEAR7 YEAR8 YEA
Slide 36: Summary
Summary PDF Solutions has a rich history in pioneering yield learning research methodologies starting form the early 80’s at Carnegie Mellon We have successfully responded to the semiconductor industry challenges for the last 25 years PDF Solutions’ Template™ layout creates a new interface bet
Slide 37: Thank You
THANK YOU
Slide 38: Products & Solutions
Products & Solutions Kimon Michaels, Ph.D. Vice President Products & Solutions This presentation contains forward-looking statements regarding PDF Solutions’ future products and business prospects that involve risk and uncertainty. Actual results could differ materially from those discussed. You sho
Slide 39: Characterization Vehicle Infrastructure (CVi)
Characterization Vehicle® Infrastructure (CVi™) CV® Solution pdFasTest® Tester Exensio® Analysis - Rich design of - Massively parallel - Automated analysis experiments test - Product yield impact - Product relevant - Optimized for CV® - Root cause - Patented structures testing diagnosis Up to 30x de
Slide 40: CVi Breadth
CVi™ Breadth Implemented in variety of vehicles • Process development test chips for foundry • IP validation test chips for fabless • Scribelines for mass production reticles Incorporate product-relevant content Scribe CV® DirectProbe™ for mass production CV® derived from fabless designs and IP
Slide 41: CVi Momentum in Electrical Characterization
CVi™ Momentum in Electrical Characterization PDF Solutions now generates ~2 CV® tapeouts per week • > 10 nodes characterized over past year CVi™ technology in use at > 30 foundry-process node combinations >100B measurements in past year Indicative of growing need for electrical characterizat
Slide 42: Competitive Advantage of CVi
Competitive Advantage of CVi™ Advantage CVi™ versus conventional infrastructure Coverage & throughput • Up to >1000x improvement in data per cm2 per unit time Product relevance • CV® results validated as representative of production yield over 100s of CV® test chips and 10,000s of wafers Fabless con
Slide 43: PDF Solutions Business Model Drives Unique Development
PDF Solutions Business Model Drives Unique Development pdFasTest® Development Incentivized by achieving results for clients Provides access and understanding of our unique data Enables targeted Latest generation CVi™ investment in ways solution is >1000x faster vs. conventional architectures c
Slide 44: PDF Solutions CV Ecosystem Expansion
PDF Solutions CV® Ecosystem Expansion Contact Probed E-test E-beam Probed E-test pdFasTest® eProbe™ E-Test tool E-beam tool On-wafer CV® 0 1 2 3 4 5 6 8 9 1011121314 16 17 1819202122 24 25 2627282930 32 33 3435363738 31 31 31 tester (array, In-chip CV® 40 41 4243444546 48 49 5051525354 56 57 5859606
Slide 45: Design-for-Inspection (DFI)
Design-for-Inspection™ (DFI™) Process Tool DFI™ Design Framework FDC and Inspection R2R Control Metrology & Review DFI™ Scribe, DFI ™ Test Chips Filler Cells Billions of Exensio™ DFI Cells Big Data Analysis Product Chip or per wafer MPW Block eProbe™ FIRE™ Layout Analysis Measurement Tool Fabless em
Slide 46: Exensio Big Data & Analytics Is Instrumental to CVi/DFI
Exensio® Big Data & Analytics Is Instrumental to CVi™/DFI™ CV®/ One CV®/DFI™ lot (25 wafers) can Process/Metrology Tool DFI™ have >1TB of associated data • Up to 10B electrical measurements per wafer • Linked to data from an alphabet soup of enterprise systems – MES; EES; FMS; RMS, TDS, etc. Exe
Slide 47: Exensio Platform – Connecting the supply chain
Exensio® Platform – Connecting the supply chain Fab Sort Final Test Fully integrated leading Yield Management (YMS) and Process Control systems 90 companies in 20 countries Industry leadership • Exensio® -Yield #1 in foundry and fabless • Exensio® -Control #1 in foundry • Exensio® -Test #2 in
Slide 48: Template Layout: Solve Design-for-Manufacturability
Template™ Layout: Solve Design-for-Manufacturability Conventional design IP approach is too Unconstrained designer iterative and slow for advanced nodes “creativity” hurts time to market 1. Define design kit that attempts to restrict unmanufacturable patterns 2. Design product / IP, fabricate, the
Slide 49: Template Technology: Solve Design-for-Manufacturability
Template™ Technology: Solve Design-for-Manufacturability Template™ Pattern Pattern B Template™ layout enables first-time-right Yield design with set of trusted patterns for IP design X Process window Process window 1. Identify trusted set of highly manufacturable patterns from breadth of CV® data
Slide 50: PDF Solutions Advantage: Design – Process Co-Optimization
PDF Solutions Advantage: Design – Process Co-Optimization Design IP & analytics Fabless Template™ & Fire™ Foundry Big data analytics Electrical Exensio® characterization CVi™ & DFI™ PDF Solutions products address the process-layout interaction dominating time to volume and necessary for mass produ
Slide 51: Solutions Provided
Solutions Provided PDF Solutions integrates our products into solutions that enable optimization of process, design and fabrication of semiconductor devices for high yield, low cost and high performance Solution Provided Client Benefit Products incorporated Integrated Yield Ramp • Reduce time to m
Slide 52: PDF Solutions Co-optimizes Fabless & Foundry Needs
PDF Solutions Co-optimizes Fabless & Foundry Needs Fabless 1 CV® for tech development layout Fabless 2 Fire™ CVi™, DFI™ with fabless metrics IP analytics creation Fabless 3 IP Tech development CVi™ incorporates content from multiple lead fabless • Benefit: Development targets and validates fabless
Slide 53: PDF Solutions Differentiation & Sustainability
PDF Solutions Differentiation & Sustainability Established executive level relationships 1997 and business with both foundries and fabless >15 years defining & servicing interface between fabless and foundry Statistics for Past 4 Years Established solutions business model CV® Scribe CV® sets >
Slide 54: Thank You
THANK YOU
Slide 55: Integrated Yield Ramp
Integrated Yield Ramp Cees Hartgring, Ph.D. VP Client Services & Sales This presentation contains forward-looking statements regarding PDF Solutions’ future products and business prospects that involve risk and uncertainty. Actual results could differ materially from those discussed. You should revi
Slide 56: Integrated yield ramp (IYR)
Integrated yield ramp (“IYR”) Yield is ramped up by reducing the defect density on a wafer • From high defect density to asymptotically approaching a “mature” defect density Industry average is ~ 7% reduction per month of the gap to that mature defect density (without PDF Solutions engagement)
Slide 57: Yield ramp curve
Yield ramp curve Yield curve for 7% reduction per month PROCESS LIFE CYCLE 100% 100 TECHNOLOGY MASS MANUFACTURING YIELD RAMP YIELD 90% DEVELOPMENT 90 VOLUME 80% 80 70% 70 60% 60 50% WAFER 50 VOLUME 40% 40 30% 30 20% 20 10% 10 0% 0 YEAR1 YEAR2 YEAR3 YEAR4 YEAR5 TIME YEAR6 YEAR7 YEAR8 YEAR9 YEAR10 5
Slide 58: Yield ramp curves
Yield ramp curves Yield curves for 7% and 14% reduction per month PROCESS LIFE CYCLE 100% 100 TECHNOLOGY MASS MANUFACTURING YIELD RAMP YIELD 90% DEVELOPMENT 90 VOLUME 80% 80 70% 70 60% 60 50% WAFER 50 VOLUME 40% 40 30% 30 20% 20 10% 10 0% 0 YEAR1 YEAR2 YEAR3 YEAR4 YEAR5 TIME YEAR6 YEAR7 YEAR8 YEAR
Slide 59: Value of an integrated yield ramp: fixed wafer starts
Value of an integrated yield ramp: fixed wafer starts Higher revenues per wafer in first 5 years of production, if die priced • Typically 5-10% on average, $0.5B net margin increase for a 40K WSPM fab PROCESS LIFE CYCLE 100% 100 TECHNOLOGY MASS MANUFACTURING YIELD RAMP YIELD 90% DEVELOPMENT 90 VOL
Slide 60: Value of an integrated yield ramp: pull in wafer starts
Value of an integrated yield ramp: pull in wafer starts Designed to get to volume faster by ~ 1.5 years (or more) • Example increase in revenues at 1.5 years pull-in for a 40K WSPM fab ~ $2B PROCESS LIFE CYCLE 100% 100 TECHNOLOGY MASS MANUFACTURING YIELD RAMP YIELD 90% DEVELOPMENT 90 VOLUME 80% 80
Slide 61: PDF Solutions' IYR business model
PDF Solutions’ IYR business model Value from a PDF Solutions’ IYR depends on level of acceleration realized and volume of wafers produced. Consequently PDF Solutions’ IYR business model is: A “Gainshare” royalty per wafer produced • Royalty depends on the results achieved and targeted to represe
Slide 62: How the IYR has evolved
How the IYR has evolved Most IYR engagements now include services in the Technology Development and Mass Manufacturing phases PROCESS LIFE CYCLE 100% 100 TECHNOLOGY MASS MANUFACTURING YIELD RAMP YIELD 90% DEVELOPMENT 90 VOLUME 80% 80 70% 70 60% 60 50% WAFER 50 VOLUME 40% 40 30% 30 20% 20 10% 10 0%
Slide 63: Example - early mass production service extension
Example - early mass production service extension Fab suffered from variability in the 4.5% manufacturing line resulting in 13% of 4.0% wafers produced being “low-yielding”. TIGHT DISTRIBUTION 3.5% (5% TAIL) PDF Solutions deployed Exensio® platform, 3.0% trained about 100 Equipment and Process 2
Slide 64: How does PDF Solutions accelerate a yield ramp?
How does PDF Solutions accelerate a yield ramp? Semiconductor manufacturing consist of many processing steps. Yield ramps rely largely on trying out different process recipes for processing steps on a product or test chip, to find out which ones work and which ones don’t – these are called “Lear
Slide 65: Competition
Competition PDF Solutions’ technology is “integrated” as one system. This makes it harder for “single element” suppliers like tester companies, design companies or software companies, to break into its market PDF Solutions’ technology benefits from many cycles of learning. Therefore, it’s extrem
Slide 66: Market trends
Market trends Total Semiconductor market continues to grow Pure-Play Foundry Sales by Feature Size ($B) 1994-2004 CAGR 7.7% $20 2015 2016F 2004-2014 CAGR 4.5% 2014-2019 CAGR (Forecast) 4.9% $15 Source: IC Insights Growth is in the latest nodes $10 TAM for PDF Solutions’ IYR: $5 ~ 1% of latest
Slide 67: Market trends
Market trends Semiconductor Industry continues to consolidate into foundries • Need volume of multiple design companies to get return on investment in next node • Less and less IDMs will be able to afford own manufacturing lines This represents growth opportunity for PDF Solutions’ IYR business:
Slide 68: Market trends - Fabless IC companies
Market trends - Fabless IC companies The manufacturing processes are becoming increasingly complex: • The fabless resources required to bring up a product in a new process are becoming huge • Foundries are less and less able to duplicate process of the “Leading Foundry” and to provide a “second so
Slide 69: Summary
Summary PDF Solutions’ unique and proprietary ecosystem of electrical characterization drives increasingly larger value at its clients 25 year history of creating electric characterization CV® test structures and accelerating yield ramps Proven business model and customer benefit 69 / EXTERNAL
Slide 70: Thank You
THANK YOU
Slide 71: Design-for-Inspection
Design-for-Inspection™ Dennis Ciplickas, Ph.D. Sr. Director of Engineering This presentation contains forward-looking statements regarding PDF Solutions’ future products and business prospects that involve risk and uncertainty. Actual results could differ materially from those discussed. You should
Slide 72: Why Is A New In-line Process Control Paradigm Needed?
Why Is A New In-line Process Control Paradigm Needed? Defects in advanced manufacturing Major yield issues are technologies are driven by interaction systematic, not random between extremely tight process windows and manufacturing variations Defects are becoming 3-D in nature, often Must eva
Slide 73: What Is Design-for-Inspection (DFI)?
What Is “Design-for-Inspection™ (DFI™)?” DFI™ exploits the unused space inherent in any System-on- Chip (SoC) logic design • Non-functional “Filler cells” commonly occupy 5-30% of logic blocks simply to ensure all transistors can be wired together E-beam enabled DFI™ Filler Cells are designed to
Slide 74: DFI Fill Statistics and Why The eProbe System Is Needed
DFI Fill Statistics and Why The eProbe® System Is Needed 7B DFI filler cells/wafer (~6% of total die area) • >250 die/wafer • >30M filler cells per die Standard E-beam tools need 10s of hours to scan each die 1000s of hours for full wafer • Conventional E-beam imaging measures all pixels in a
Slide 75: How Does DFI Work?
How Does DFI Work? Measure electrical DFI cells are designed to response index of DFI catch key process issues, pad using e-beam tool leveraging PDF Solutions’ Key pattern of interest proven Characterization (in this case, V0 short Vehicle® infrastructure to CA) • Technology-specific hard IP Embed
Slide 76: DFI is a Complete System
DFI is a Complete System Process Tool DFI™ Design Framework FDC and Inspection R2R Control Metrology & Review DFI Scribe, DFI Filler Test Chips Cells Billions of Exensio® DFI™ Cells Big Data Analysis Product Chip or per wafer MPW Block eProbe® FIRE™ Layout Analysis Measurement Tool DFI combines ha
Slide 77: Competitive Advantages of DFI
Competitive Advantages of DFI Conventional E-beam Hot-spot Optical Advantage DFI E-beam Inspection (incl. predictive) Inspection Coverage Billions of DFI cells/hr • mm2/hr, tiny pixel • 100Ks hotspots/hr Full wafer and Large E-beam spot size • Must choose • Must choose Throughput limited care areas
Slide 78: DFI Deployment Status
DFI Deployment Status Fabless Product or MPW Logic Total tapeouts with DFI Test Chip tapeouts with DFI 25 + 4 in progress 60 Content Fabless companies running chips 50 5 + 2 in progress with DFI content 40 IP Vendors licensed to distribute 1 + 1 in progress DFI Fill Cells 30 CV® Tapeouts with DFI co
Slide 79: DFI Gaining Traction
DFI Gaining Traction Two eProbe® -150 systems deployed in 2016 and in active use • System #1 up and running for several months, producing results on daily basis • System #2 shipped last quarter, installed and produced results in less than 6wks, currently integrating into fab automation systems for
Slide 80: DFI Application Areas and Target Markets
DFI Application Areas and Target Markets IC Product design Ramp Full production Process development IP and Process Characterization Scribeline Wafer Health Mass Production Control DFI contributes to the entire lifecycle of an advanced logic technology • Major focus is 28nm and below • DFI helps ra
Slide 81: DFI Business Model
DFI Business Model eProbe® -150 systems deployed under two business models • System #1 deployed as part of IYR engagement as additional technology • System #2 deployed to non-IYR customer on a standalone basis Business Model for eProbe® -150 – IYR deployment • eProbe® -150 system a component of
Slide 82: Where Are We In The DFI Journey?
Where Are We In The DFI Journey? Accomplishments Hurdles remaining DFI system flow proven to work with test Ramp production of eProbe® -150 chips and is actively being used for yield learning in leading edge logic nodes Deploy multiple eProbe® -150 systems per fab eProbe® -150 and Exensio® -
Slide 83: DFI Positioned For Success
DFI Positioned For Success Next generation DFI is a natural extension of PDF pdFasTest-F Capability, Throughput Solutions’ core business: pdFasTest-F 3xx efficient characterization based on pdFasTest-F custom IP, measurement hardware, and 1xx/F2xx analysis systems pdFasTest-III pdFasTest-II E-test
Slide 84: DFI Summary
DFI Summary New fab start-up and the inline characterization and metrology of complex new processes needs to be simpler and more effective DFI offers unique E-beam capability for fast, inline, electrically relevant characterization, cutting 6wks from typical product wafer learning cycles DFI p
Slide 85: Thank You
THANK YOU
Slide 86: Exensio: Platform for big data analytics and control
Exensio ® Platform for big data analytics and control Said Akar Ph.D. GM Volume Manufacturing Solutions Group This presentation contains forward-looking statements regarding PDF Solutions’ future products and business prospects that involve risk and uncertainty. Actual results could differ materiall
Slide 87: Exensio: Real-time control for fab and assembly floor
Exensio® Real-time control for fab and assembly floor Process and Test floor product monitoring characterization and control Engineering data Analysis fab to test floor 87 / EXTERNAL USE © 2016 PDF Solutions, Inc. All rights reserved.
Slide 88: Exensio - Ecosystem build out
Exensio - Ecosystem build out Deals Closed (Normalized) Driven by integration of 3.5 process control and yield data in the fab; the 3.0 introduction of new 2.5 capabilities for test; Software as a Service 2.0 (SaaS) 1.5 1.0 Industry taking advantage of a larger Exensio 0.5 ecosystem 88 / EXTERNA
Slide 89: Exensio business model
Exensio business model Licensing methods: • Revenue primarily ratable with a few historic perpetual contracts • Majority of licenses are time based • Licenses for Exensio -Control and Exensio -Test are tied to number of tools connected and scales with volume • Licenses for Exensio –Char scale with
Slide 90: Fabless – Fabs – Foundries – OSATs – System Houses
Fabless – Fabs – Foundries – OSATs – System Houses 90 companies in 20 countries >21K process tools under PDF process control Exensio® –Yield #1 in foundry Exensio® –Control #1 in foundry Exensio® –Test #2 in fabless/fab-lite PDF Headquarters PDF Support/Development Centers Representative i
Slide 91: Exensio – Connecting the value chain
Exensio – Connecting the value chain Fab Sort Final Test Fully integrated leading Yield Management (YMS) and Process Control systems • Strategic use of proven Big Data open source software components for cost and performance • Integration of world-class analytics and visualization • Complete integ
Slide 92: Exensio - The integrated system
Exensio - The integrated system Interconnected and complementary objectives Test Better Test and Assembly Ops drives quality Quality Better Characterization drives quality and yield Yield Better Yield and Quality Link drives NPI Time-to-Market Link Better Control and Strategy at test and a
Slide 93: Manufacturing transformation – at OSAT
Manufacturing transformation – at OSAT‡ “A core element of Mfg. transformation program is the work with PDF Solutions… Big Data – Systems ready for the data tsunami” Michael Chang, VP Test Operations Presented by Michael Chang, Vice President of test operation of SPIL at PUG2016 – Used with Permissi
Slide 94: Test operations and yield – Exensio at Fairchild
Test operations and yield – Exensio at Fairchild “First-pass yield improvement of 10-20%; Tester throughput improvement of 30%” Dr. Wei-Chung Wang, SVP Manufacturing Operations http://www.pdf.com/press-releases/Fairchild_Semiconductor_Adopts_PDF_Solutions_Exensio__-Test_Module__For_Global_Test_Opera
Slide 95: Semiconductor market size for Exensio
Semiconductor market size for Exensio Real-Time Fab Offline Fab Offline Product Real-Time product Analysis Analysis Analysis Analysis $100M $200M The initial market for Exensio is the external served market is ~$150M, much served by smaller private Exensio Functional Coverage companies As customers
Slide 96: Internal Solutions
Internal Solutions FDC / APC Lack of integration - TOOL Multiple Databases, FDC METROLOGY Multiple User Interfaces FAB Team COMSUMABLES ENVIRONMENTAL YMS Silos with only part of the MAINTENANCE picture available to each SCHEDULES (PM) DESIGN Team team/sub-team PCM/SCRIBE Defect CV TESTCHIP Min
Slide 97: Data growth rate in manufacturing
Data growth rate in manufacturing Required data for successful ramp, at each subsequent node, grows at exponential rate. Main growth is typically in characterization and manufacturing data. Data sizes still lag order of magnitudes behind other industry applications of Big Data. Typical Foundry
Slide 98: Opportunities - big data architecture
Opportunities - big data architecture PDF Solutions applies big Message Bus (HornetQ) data technologies, originally developed to handle internet Parallel Cluster scale, to IC manufacturing Summaries Online Analysis and PDF solutions links together All Data Types Inbound API Meta Data the huge va
Slide 99: Opportunities - big data analytics
Opportunities - big data analytics If Netflix recommendations just used statistics you’d only get to see block busters and academy award winners Searching Petabyte IC manufacturing databases for actionable yield signals needs to be equally smart • Only the hard-to-find problems remain – low sign
Slide 100: Opportunities - business model extension - SaaS
Opportunities - business model extension - SaaS Cloud-based access using Software-as-a-Service (SaaS) model following Syntricity acquisition Leveraging the SaaS customer base and knowledge acquired with Syntricity to roll out Exensio Facilitating • Collaboration across value chain players • Re
Slide 101: Opportunities - assembly
Opportunities - assembly Traditionally less automated, less complexity System OEM requirements driving complexity and automation growth Forecasted assembly capital investment is outstripping both industry and fab growth* • Assembly using front-end type etch, deposition processes Assembly • Aut
Slide 102: Opportunities - test
Opportunities - test Real time collaboration between Product Engineering and Test Engineering teams OSAT and fabless IC design team collaboration opens major market Wafer Sort Final Test • Operations • Operations Control Control • Outlier Detection • Outlier Detection • OEE† • Root cause ID • OE
Slide 103: Opportunity for the entire systems value chain
Opportunity for the entire systems value chain IC Design Fab Sort Assembly Test System Design Image - Atropos235 at English Wikipedia – 103 / EXTERNAL USE © 2016 PDF Solutions, Inc. All rights reserved. CC BY-SA 2.5, https://commons.wikimedia.org/w/index.php?curid=15501313
Slide 104: Exensio competitive position
Exensio competitive position Only PDF Solutions serves the Only PDF Solutions is a major user entire manufacturing value chain of its own software and a major contributor to its evolution from a Only PDF Solutions can deliver its user perspective unique data types – CV®, scribe CV®, DFI™ Str
Slide 105: Thank You
Allders by Process Step colored ly Lot vex coer by 205, Buasre6 4 1 1 1 [Gv2dea6.1 ‘Sumi(otal_adder count (sum) + Dep Netalt Ase oo. Fe 2M ADC ‘y29_BM_Abe 16. Metalt_ADC ars 25 SRE_Dep (4 Metet_Dep Io6_Netal_ADe ‘G48 Natal _ADC 64 Metaa_ase (064 Natalt ASC or Met (0 Metas_ASC ns ‘WA0_Eni Py Dep ir e
Slide 106: China Opportunities
China Opportunities Hua Fang Ph.D. Vice President and General Manager, PDF Solutions (Shanghai) This presentation contains forward-looking statements regarding PDF’s future products and business prospects that involve risk and uncertainty. Actual results could differ materially from those discussed.
Slide 107: Semiconductor Growth is in China
Semiconductor Growth is in China +19.7% +16.1% Global China Global China -2% -3% 2015 Semiconductor Sales 2016 1H Semiconductor Sales 107 / EXTERNAL USE © 2016 PDF Solutions, Inc. All rights reserved.
Slide 108: Into China – Business Logic in 2006
Into China – Business Logic in 2006 PDF would become a big data company in Semiconductor. We need to establish an “analysis factory” to standardize analysis and benchmark. China market would be big. We need to prepare our talents well ahead of the curve. 108 / EXTERNAL USE © 2016 PDF Solutio
Slide 109: Solid Start
Solid Start 109 / EXTERNAL USE © 2016 PDF Solutions, Inc. All rights reserved.
Slide 110: Strong Growth for Ten Years
Strong Growth for Ten Years BD & Sales PDF China Headcount Trend 200 SW SW Back- Application Office 150 Engineer Support 100 Field Consultant Data Mining Application Engineer Engineer 50 Data Data Operation 0 Analysis Integration CV Design Enhancement 06 07 08 09 10 11 12 13 14 15 16 Engineer Engine
Slide 111: 2014 – IC Industry Became National Priority
2014 – IC Industry Became National Priority Tax incentive, talent attraction, Market driven to accelerate direct funding on projects scale of proven companies 111 / EXTERNAL USE © 2016 PDF Solutions, Inc. All rights reserved.
Slide 112: Establishment of US$20B+ CICIIF (Big Fund)
Establishment of US$20B+ CICIIF* (aka “Big Fund”) First round: US$20B Promote investment in IC through other channels to reach >US$100B Investment targets: leaders in each segment Investment segments: manufacturing, fabless, assembly&test, equipment, material. *CICIIF: China IC Industry Inve
Slide 113: Big Fund Investment Status
“Big Fund” Investment Status Invested in 37 projects from 29 companies Committed investment totaled US$10B • 60% in manufacturing, 27% in design, 8% in A&T,3% in equipment, 2% in material (YMTC) Design Manufacturing Equipment Assembly Fund/Eco-System 113 / EXTERNAL USE © 2016 PDF Solutions, Inc.
Slide 114: Major Activities from Other Funds
Major Activities from Other Funds Hua Capital OmniVision, ISSI Summitview Captial ISSI, ASR JAC Captial NXP RF/PA/Bipolar CGP Investment Silex, SMIC NSIG SOITEC 114 / EXTERNAL USE © 2016 PDF Solutions, Inc. All rights reserved.
Slide 115: How to Ride the Wave – Business Logic in 2014
How to Ride the Wave – Business Logic in 2014 China needs to develop technology differently in order to catch up world leaders. We need to push our most advanced technologies aggressively and design our solutions differently. In both foundry and fabless spaces, there are large gaps between lea
Slide 116: Dr. Gerald Yin Joins PDF China Board in 2015
Dr. Gerald Yin Joins PDF China Board in 2015 116 / EXTERNAL USE © 2016 PDF Solutions, Inc. All rights reserved.
Slide 117: Leaders: Rapid Growth and Continuing Aggressive Investment
Leaders: Rapid Growth and Continuing Aggressive Investment SMIC • Target to grow 5x in 5 years • Aggressive capacity expansion in Beijing (12”), Shanghai (12”), Tianjin (8”), Shenzhen (8”) and Ningbo (12” or 8”) as well as global acquisition (L-Foundry) • Aggressive investment in eco-system (fable
Slide 118: 10+ New 12" Fabs under Construction or Planned
10+ New 12” Fabs under Construction or Planned SMNC Fab3 Jinghe YMTC Tacoma GlobalFoundries TSMC SMIC SN1 & Huali AOS Fab2 UMCX & JingHua QHZG 118 / EXTERNAL USE © 2016 PDF Solutions, Inc. All rights reserved.
Slide 119: Fast Growth of Fabless IC Companies
Fast Growth of Fabless IC Companies 2015 2016 Sale Area Sales Sales Increase Company Ratio Company Ratio % (US$B) (US$B) 1 Communication 157 21.3% 9.0 241 17.7% 10.4 15.2% 2 Consumer 113 15.4% 2.2 589 43.2% 4.7 116.0% 3 Multimedia 93 12.6% 1.3 43 3.2% 2.7 109.6% 4 Smartcard 39 5.3% 1.4 69 5.1% 2.0 3
Slide 120: Center of OSATs Moving to China
Center of OSATs Moving to China JCET STATSChipPAC (100%) FTME AMD facilities in Suzhou/China and Malaysia (85%) Tzinghua Unigroup ChipMos (25%) SMIC becomes largest share- holder of JCET XMC strategic partner w/ Tianhui More aggressive investment in both capacity and technology exp
Slide 121: 2016 – Business Logic Updated
2016 – Business Logic Updated PDF has established its strategic importance with top leaders. Investments are focused on top leaders as well. We should be able to leverage the strength of the funds to increase our impact on the leaders and improve the economic returns on the impact. Promoting “
Slide 122: Thank You
THANK YOU
Slide 123: Analyst Day – Greg Walker, CFO
Analyst Day Greg Walker CFO This presentation contains forward-looking statements regarding PDF Solutions’ future products and business prospects that involve risk and uncertainty. Actual results could differ materially from those discussed. You should review PDF Solutions’ SEC filings, including it
Slide 124: Business opportunity & financial summary
Business opportunity & financial summary Evolution of PDF Solutions’ Financial Model Pre 2014 IYR Centric Financial Model 2014-2017 Investment Centric Financial Model New Investments & Early Results Post Investment Financial Model Equity & Cash Priorities Appendix - Historical Financia
Slide 125: PDF Solutions financial model: Pre-2014 model
PDF Solutions financial model: Pre-2014 model Historical financial model has evolved to address its changing business opportunities: Originally a classic technical software model (similar to EDA) Moved to a customer aligned solutions model through the early 2000’s to 2014 Project based yield r
Slide 126: PDF Solutions: IYR Centric Financial Model
PDF Solutions: IYR Centric Financial Model Pre-2014 2013 Non-GAAP in Percents of Revenue Model Actual Solutions Revenue 60% 61% Gainshare Revenue 40% 39% HOWEVER: Customer concentration and Total Revenue 100% 100% reliance on IYR business Solutions Cost of Sales 35% 36% contributed to risk levels Ga
Slide 127: Financial model: 2014-2017 investment model
Financial model: 2014-2017 investment model To broaden its market opportunities and lower its business risk the Company began a series of investments: To fund our key investments we have temporarily altered our financial model Key Investments: • Expanding our China operation to match our local c
Slide 128: Market opportunity for IYR continues to grow
Market opportunity for IYR continues to grow Total Semiconductor market continues to grow Pure-Play Foundry Sales by Feature Size ($B) 1994-2004 CAGR 7.7% $20 2004-2014 CAGR 4.5% 2015 2016F 2014-2019 CAGR (Forecast) 4.9% Source: IC Insights $15 Growth is in the latest nodes High level of Custo
Slide 129: 10+ New 12" Fabs under construction or planned
10+ New 12” Fabs under construction or planned SMNC Fab3 Jinghe YMTC China is putting 60% of its investment into fabs Tacoma GlobalFoundries Getting a fab up fast is incredibly important TSMC PDF Solutions, with investment in China and SMIC SN1 & Huali AOS technology, is well positioned to serve thi
Slide 130: Technology is driving need for new kinds of inspection
Technology is driving need for new kinds of inspection: Existing e-beam inspection market ~$270M We believe about >50% of the e-beam market is for voltage contrast We believe DFI™ with the eProbe™ 150 can better serve VC use case eProbe™ 250 can grow the e-beam market because of new use case
Slide 131: Semiconductor market size for Exensio
Semiconductor market size for Exensio Real-Time Fab Offline Fab Offline Product Real-Time product Analysis Analysis Analysis Analysis $100M The initial market for Exensio$200M is the external served market which is ~$150M, much served by smaller Exensio Functional Coverage private companies As custo
Slide 132: PDF Solutions TAM Summary $500M - $1300M
PDF Solutions TAM Summary $500M - $1300M Integrated Yield Ramp $200M-$400M • 2015 Foundry Revenue approximately $50B • Estimate foundry R&D $4.0B, roughly 50% process development, up to 20% open to PDF Solutions • Low-end of range assumes no capture of major new IYR customers (TSMC & INTC) • China
Slide 133: Declining customer concentration
Declining customer concentration 133 / EXTERNAL USE ©(c© 2016 PDF Solutions, Inc. All rights reserved.
Slide 134: PDF Solutions: investment financial model
PDF Solutions: investment financial model 2014-2017 YTD 2016 Non-GAAP in Percents of Revenue Model Actual Solutions Revenue 65% 73% Gainshare Revenue 35% 27% Making good progress against target model Total Revenue 100% 100% Solutions Cost of Sales 35% 36% Gainshare Cost of Sales 0% 0% Total Cost of
Slide 135: PDF Solutions' financial model: Post 2017 long term model
PDF Solutions’ financial model: Post 2017 long term model Post investment financial model: Growth accelerates beyond industry growth rates based on significantly enlarged TAM Revenue mix shifts as Exensio® and DFI™ become more significant Gainshare revenues increase, eventually exceeding pre-2
Slide 136: PDF Solutions: Long Term Financial Model
PDF Solutions: Long Term Financial Model 2014-2017 YTD 2016 Long Term Non-GAAP in Percents of Revenue Model Actual Model Solutions Revenue 65% 73% 75% Gainshare Revenue 35% 27% 25% Total Revenue 100% 100% 100% Solutions Cost of Sales 35% 36% 25% Gainshare Cost of Sales 0% 0% 0% Total Cost of Sales 3
Slide 137: PDF Solutions: Financial Model & Investor Interests
PDF Solutions: Financial Model & Investor Interests PDF Solutions’ financial model well aligned with investor interests: High leverage for profits in financial model Tight cost controls and selective investments All bonus programs and executive compensation tied directly to achieving profitabi
Slide 138: PDF Solutions: Equity and cash investment strategies
PDF Solutions: Equity and cash investment strategies Cash investment prioritization: Business Investments (Internal Spending/ 3rd Party/ M&A) a. DFI™ success and acceleration b. Exensio® capability expansion c. China expansion d. Improved operational efficiency (e.g. Netsuite/Salesforce) Anti-di
Slide 139: PDF Solutions financial model summary
PDF Solutions financial model summary Investing in growth and expanding TAM Broadening technologies, markets and customers Positioned for China investment and growth Developing new technology for Process Control Early adopter of Web data technology for MFG 4.0 (Exensio®) Capturing revenu
Slide 140: Appendix - PDF Solutions financial historical results
Appendix - PDF Solutions financial historical results Income statement Royalty revenues Margins Cash 140 / EXTERNAL USE ©(c© 2016 PDF Solutions, Inc. All rights reserved.
Slide 141: PDF Solutions: Non-GAAP P&L
PDF Solutions: Non-GAAP P&L P&L Non-GAAP 2013 2014 2015 2016 YTD Design to Silicon Yield Solutions 61,710 52,769 64,741 57,872 Gain Share - Performance Incentive 39,743 47,394 34,138 21,324 Total Revenue 101,453 100,163 98,878 79,196 Cost of Sales 36,735 34,403 36,825 28,802 GROSS MARGIN 64,718 65,7
Slide 142: Cumulative royalty revenue from 2011-2015 ($M)
Cumulative royalty revenue from 2011-2015 ($M) Source: Company Reports 142 / EXTERNAL USE ©(c© 2016 PDF Solutions, Inc. All rights reserved.
Slide 143: PDF Solutions: historical financials – margins (Non-GAAP)
PDF Solutions: historical financials – margins (Non-GAAP) Gross Margin % (non-GAAP) Op Margin % (non-GAAP) 80% 70% 60% 50% 40% 30% 20% 10% 0% 2015 & 2016 Operating Margins reflect significant increase in R&D spending related to DFI™ development expenses 143 / EXTERNAL USE ©(c© 2016 PDF Solutions, In
Slide 144: PDF Solutions: cash balance
PDF Solutions: cash balance $000 Cash & Investments 140,000 120,000 100,000 80,000 60,000 40,000 20,000 0 Q1-12 Q2-12 Q3-12 Q4-12 Q1-13 Q2-13 Q3-13 Q4-13 Q1-14 Q2-14 Q3-14 Q4-14 Q1-15 Q2-15 Q3-15 Q4-15 Q1-16 Q2-16 Q3-16 In 2015 the Company spent $14.5M on stock repurchases under a Board approved p
Slide 145: Thank You
THANK YOU