🛠️ Career Summary
Shinichi Samizo is an engineer who has continuously deepened his expertise from
control theory & electromagnetic analysis, through semiconductor device development
and PZT actuators, to the commercialization of PrecisionCore printheads —
evolving from technology → systems → education.
📘 Career Phases
⚡ Phase 0|Control Design & Electromagnetic Analysis (1994–1997)
-
Experienced control design using MATLAB / Simulink as an undergraduate.
📘 Digital H∞ Control -
Graduate research in electromagnetic field analysis.
Episode: Thin-Film Microreactor Analysis (1996–1997)
Analyzed DC-DC converter reactors using ferrite-based magnetic materials and Al spiral coil structures.
Presented guidelines for Q-factor degradation and Al/Cu material selection at 500kHz–1MHz.
🔬 Thin-Film Microreactor 1996
💾 Phase 1|Semiconductor Device Development (1997–2006)
-
1997: Joined Seiko Epson Corporation.
Engaged in logic, memory, and high-voltage process integration (0.35–0.18μm generations). -
DRAM/VSRAM Technology Archive
Participated in ramp-up of 0.25µm DRAM (1998) and VSRAM (2001).
Led process setup, defect analysis, and yield improvement.
📎 DRAM/VSRAM Technology Archive
Documentation of ramp-up activities for DRAM and VSRAM, including process setup methods, case studies of defect analysis, and yield improvement practices. -
Developed high-voltage embedded technology (30V transistors), applied to a-TFT driver ICs.
- ⚙️ 0.18μm CMOS Embedded Front-End Process
Developed high-voltage embedded front-end technology (30V devices) integrated into standard CMOS flow, applied to a-TFT driver ICs. - 📑 LCD Driver IC Specification
Specification for LCD Panel Driver IC (concept model: power rails, I/Os, gamma reference, temperature compensation, timing, pixel circuit).
- ⚙️ 0.18μm CMOS Embedded Front-End Process
🎛️ Phase 2|Piezoelectric Materials & Actuators (2007–2012)
-
Started with PZT process evaluation for FeRAM, later applied to thin-film piezo actuators.
🔧 0.18μm FeRAM Process Flow -
Contributed to structural analysis and reliability evaluation of PZT films, establishing the foundation of PrecisionCore technology.
-
Overcame thin-film reliability issues through defect analysis and surface treatment optimization.
🖨️ Phase 3|PrecisionCore Commercialization & Education (2012–)
-
Led COF implementation and driver IC integration for PrecisionCore printheads.
🔥 SystemDK: Thermal / Stress / Noise Constraints -
Responsible for scheduling and execution from development to mass production.
-
Promoted BOM management, ISO training, and knowledge transfer, transforming them into structured educational content.
📑 Design-to-Mass Production Ordering Flow
🎯 Current Activities
-
Building and sharing Samizo-AITL, an open framework for education, prompt design, and AI-integrated control.
-
Publishing structured educational resources integrating experiences from semiconductors, control, and inkjet.
📄 FinFET / GAA Node Parameter Comparison
📑 Papers & Research Works
🧩 Historical Case Study: 0.25-µm DRAM & VSRAM
Case study of 0.25 µm DRAM ramp-up and VSRAM mass production.
Analyzes process transfer at Epson’s Sakata Fab, yield-improvement cycles, and defect analysis.
Provides historical insights into process migration and mobile memory development.
📄 DRAM & VSRAM History Paper (PDF)
🧪 Historical Case Study on Ti Silicide (TiSi₂) Reliability Issues
Reliability study of TiSi₂ phase-transition instability at 0.25 µm CMOS.
Examines incomplete C49→C54 transformation, boron absorption, and SRAM yield loss.
Highlights redundancy limitations, etch/RTA countermeasures, and scaling risks.
📄 TiSi₂ Reliability Case Paper (PDF)
🎯 Process-Based Differentiation for Analog CMOS
Process-based approach reducing MOSFET 1/f noise by >50% at 0.18 µm CMOS.
Uses epi substrate, optimized doping, oxide control, H₂ annealing, and geometry tuning.
Demonstrates robustness across temperature/aging, with impact on analog/sensor ICs.
📄 CMOS 0.18 µm Noise Reduction Paper (PDF)
🔋 Low-Cost Integration of 1.8-V FeFET on 0.18-µm CMOS
FeFET integrated into 0.18 µm CMOS with +1 mask and ALD process.
Shows endurance >10⁵ cycles and retention >10 years at 85 °C.
Enables IoT/automotive non-volatile memory for backup and secure key storage.
📄 FeFET CMOS Reliability Paper (PDF)
⚡ On-Chip Magnetic-Laminated Inductor in 0.18-µm CMOS
On-chip laminated inductor with PGS integrated into 0.18 µm CMOS.
Implemented in hybrid Buck–LDO regulator achieving >80% efficiency and wideband response.
Provides competitive alternative to external inductors for IoT and automotive SoCs.
📄 CMOS018 Inductor + LDO Paper (PDF)
🖥️ SystemDK for 3D-IC
Constraint-driven DTCO framework integrating multi-physics into EDA.
Maps FEM thermal/stress and EMI analysis into STA/placement constraints.
Validated on TSV stacks: 87% slack recovery, 11 °C hotspot reduction, 23% eye-opening gain.
📄 SystemDK for 3D-IC Paper (PDF)
💾 LPDDR+FeRAM Chiplet Integration
Hybrid memory combining low-power LPDDR with FeRAM chiplets.
SystemDK co-design enables reduced standby power and instant resume.
Applies to IoT/automotive systems requiring efficient, reliable memory.
📄 LPDDR+FeRAM Integration (PDF)
📘 Educational Tutorial Paper on CFET
Educational overview of Complementary FET (CFET) technology.
Reviews device evolution: Planar → FinFET → GAA → CFET.
Highlights design/manufacturing challenges and teaching applications.
📄 CFET Tutorial Paper (PDF)
🖥️ SystemDK with AITL Paper
Runtime-aware DTCO framework with PID+FSM+LLM control loops.
Real-time compensation of RC delay, thermal, and EMI variability.
Reduces guardbands and enhances sub-2nm reliability.
📄 SystemDK+AITL Main Paper (PDF)
⚙️ Control-Integrated CFET Study
Proof-of-concept for cross-layer CFET control.
Applies PID+FSM+LLM for delay and thermal compensation in nanoscale nodes.
Demonstrates dynamic reliability improvement in sub-2nm ICs.
📄 CFET Control Main Paper (PDF)
🤖 Humanoid Control Architecture Research
Control framework integrating LLM, FSM, PID, and state-space methods.
Validated via cross-node SoC and SystemDK co-design.
Supports next-generation humanoid robots with adaptive control.
📄 Humanoid TCST Paper (PDF)
🚀 AITL on Space
Adaptive Intelligent Triple-Layer (AITL) control with Tri-NVM hierarchy.
Combines H∞ control, FSM, and AI re-design on 22 nm FDSOI SoC.
Enables long-duration spacecraft autonomy and reliability.
📄 AITL on Space Main Paper (PDF)
✈️ SkyEdge High-Altitude Drone Platform
Reference design for secure high-altitude UAV with H∞ control.
Features sensor fusion, variable-pitch rotor, PQC security, and FSM/LLM control.
Ensures robust performance and reliable operation up to 10 km.
📄 SkyEdge Drone Paper (PDF)
🧮 Post-CFET Device Architectures
Comprehensive survey of post-CFET device architectures (2030–2045).
Compares 2D FETs, monolithic 3D, spintronics, and quantum devices.
Provides roadmap on materials, integration, reliability, and EDA/PDK needs.
📄 Post-CFET Main Paper (PDF)
🖨️ Inkjet Tutorial Paper
Comprehensive tutorial on inkjet principles, materials, and applications.
Covers actuation modes, droplet physics, ink chemistry, system/device design, and analysis.
Outlines future directions: Pb-free piezoelectrics, low-voltage actuation, and bio-integration.
📄 Inkjet Tutorial Paper (PDF)
🌊 ScAlN MEMS Ultrasonic Sensor Study
Pb-free ScAlN MEMS integrated with 65 nm SiGe CMOS via SiP.
Demonstrates eco-friendly, high-sensitivity ultrasonic MEMS arrays.
Targets next-generation medical imaging and non-destructive sensing.
📄 ScAlN Ultrasonic Paper (PDF)
🖨️ KNN Bio-Inkjet Printing Study
Bio-Inkjet system using Pb-free KNN actuators under ±50 V drive.
Achieves picoliter-scale droplet generation for biological samples.
Enables safe, sustainable bio-printing of cells and proteins.
📄 Bio-Inkjet Paper (PDF)
⚙️ SystemDK Inkjet Framework
Design support framework for industrial piezoelectric inkjet systems.
Unifies electrical, mechanical, and fluidic domains into a System Design Kit (SystemDK).
Improves droplet control accuracy, shortens design cycles, and accelerates prototyping.
📄 SystemDK Inkjet Paper (PDF)
⚠️ Note
The process information described here is based on conceptual models for educational purposes
and is not related to actual manufacturing flows or confidential data.