Kioxia Holdings Corporation · TSE:285A

Kioxia — Investor Day (June 2026)

43 slides from Kioxia's investor day presentation, Jun 2, 2026

Slide 2: Agenda
Agenda 1. Kioxia’s Growth Strategy in the AI Inference Era Hiroo Ota President and CEO 2. Flash Memory Market and Resource Allocation Junichiro Yaguchi Managing Executive Officer (CSO) 3. SSD Drives AI inferences Masashi Yokotsuka Managing Executive Officer (Vice President of SSD Division) 1 4. Adva
Slide 3: Disclaimer
Disclaimer Forward-looking statements included in this presentation are prepared based on our expectations and projections in light of the information currently available to us, which involve various risks and uncertainties (including, but not limited to, economic trends, market demand and the highl
Slide 4: Kioxia's Growth Strategy in the AI Inference Era
Kioxia’s Growth Strategy in the AI Inference Era Hiroo Ota President and CEO
Slide 5: Agenda
Agenda ◼ Expanding the Role of Flash Memory in AI Inference ◼ Focusing on the Data Center and Enterprise Markets ◼ Growth Investments to Further Enhance Corporate Value © 2026 KIOXIA Holdings Corporation. All Rights Reserved. 5
Slide 6: Review of the Past Year
Review of the Past Year Our corporate value increased to more than 30x, driven by expanding AI data center demand Our stock price trend (Yen) Stock price increase 70,000 reflecting expectations for business 60,000 performance 50,000 Rising memory prices and 40,000 tighter supply- Explosive growth de
Slide 7: Rapid Growth in Inference Workloads
Rapid Growth in Inference Workloads Agentic AI and Physical AI that autonomously solve tasks are driving inference workloads to increase exponentially Agentic AI User Usage AI using AI Urgent Need to Improve Inference Physical AI Processing Efficiency People using AI © 2026 KIOXIA Holdings Corporati
Slide 8: NVIDIA's Proposed GPU-Connected Storage
NVIDIA’s Proposed GPU-Connected Storage A concept to redefine SSDs as a memory extension layer for GPUs 推論GPUサーバ Inference GPU Server RAGサーバ RAG Server: Improving Answer Accuracy by Retaining Knowledge DB generation DB search DRAM コンテキスト Improved Inference Efficiency Expanding the GPU キャッシュサーバ Memor
Slide 9: Focus on the Data Center & Enterprise Markets
Focus on the Data Center & Enterprise Markets Kioxia Sales Portfolio Excluding SanDisk sales revenue Revenue mix of products for the Data Center & Enterprise markets Over 60 % Driving profitability through strategic product-market fit for value-added products (e.g., Super High IOPS SSDs) As Is To Be
Slide 10: Phase Change
Phase Change Two consecutive years of record-high sales revenue and operating profit Growth investments including inorganic investments Significant improvement Achieve in profitability net cash position Q1 OP Margin: 74%1 in Q1 FY26 Shareholder returns 1. FY2026 Q1 Guidance (Announced on May 15, 202
Slide 11: Growth Investments to Enhance Corporate Value
Growth Investments to Enhance Corporate Value Advancing AI Memory Solutions from Multiple Angles 3 2 Inorganic Investments 1 Turning R&D Results into Businesses ✓ Capturing Adjacent Businesses as AI Evolves Strengthening Core Operations ✓ Horizontal Channel Flash Memory (HCF) ✓ Product and Technolog
Slide 13: Flash Memory Market and Resource Allocation
Flash Memory Market and Resource Allocation Junichiro Yaguchi Managing Executive Officer (Chief Strategy Officer)
Slide 15: Flash Memory Market
Flash Memory Market Flash Memory Market (EB) Flash Memory Market for Data Centers 1,807 (EB) (EB) CAGR 22% Data Center 909 46% 997 909 CAGR Others 46% DC 295 Inference PC 86 % PC / PC / Smartphone Market Smartphone 295 Training Flat or Slightly Declining 8% 16% Smart‐ phone Traditional 16% （CY） '25
Slide 16: Supply-Demand Balance and Revenue Market Forecast
Supply-Demand Balance and Revenue Market Forecast Supply-Demand Balance Flash Memory Market by Revenue Base Tight Supply-Demand to Persist Throughout FY27 Rapid Market Expansion Driven by Rising ASPs 500 (B$) 110% 400 300 100% 200 100 90% 0 Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4 （CY） '24 '25 '26E '27E
Slide 17: Production Strategy
Production Strategy 1 Capital Investment Plan Investment Efficiency Trend Planned Annual Capex: Approx. JPY 470 billion High Investment Efficiency (billion yen) 5-Year Cumulative Capex / Annual GB Output ($/GB) 470 Industry Average KIOXIA 66% Source: TechInsights Inc. NAND Market Report Q2 2026 （CY）
Slide 18: Production Strategy: Front-End Cost per Gigabyte
Production Strategy 2 Front-End Cost per Gigabyte Cost per Gigabyte (CY25) Front-End GB Cost Reduction via Node Cost Competitiveness via Efficient Capital Transition Investment BiCS FLASH BiCS FLASH BiCS FLASH Gen.6 Gen.8 Gen.10 Bit density Bit density x1.5 x1.6 Maximizing Bit Density Beyond Layer S
Slide 19: R&D Strategy
R&D Strategy R&D Expenses Aligning R&D Pace with the AI-Driven Evolution of the Memory Hierarchy Annual R&D investment: JPY 230 billion (billion yen) 230 Performance ＆ Cost 63% DRAM OCTRAM, etc. 141.1 New Memory XL-FLASH Device Development BiCS FLASH Gen.10/Gen.11 TLC SSD development QLC New areas i
Slide 20: SSD Drives AI Inference
SSD Drives AI Inference Masashi Yokotsuka Managing Executive Officer (Vice President of SSD Division)
Slide 21: Agenda
Agenda ◼ Achievements from FY2025 ◼ Challenges Faced by Growing Inference Systems ◼ New Uses of SSD Created by Inference AI Systems ◼ Our Latest SSD Lineup to Solve Scaling Challenges in AI © 2026 KIOXIA Holdings Corporation. All Rights Reserved. 21
Slide 22: Achievements from FY2025
Achievements from FY2025 Software for Improving Inference Accuracy and Speed Ultra Large Capacity SSD KIOXIA LC9 Series KIOXIAAiSAQ Enhancing RAG Accuracy: Constructing Large-Scale ～ 245.76 TB Vector Databases on SSDs for Rapid Searches Powering Generative AI with Accelerating indexing with GPUs thr
Slide 23: Challenges Faced by Growing Inference Systems
Challenges Faced by Growing Inference Systems 推論GPUサーバ Inference GPU Server RAG Server：Improving Answer Accuracy RAGサーバ through Knowledge Retention ! ! DRAM Insufficient GPU Memory Capacity Due to the Expansion of KV Cache through Complexity in Inference - DRAM Can't Keep Up with Growing Database (D
Slide 24: New Uses for SSDs Created by the Latest Inference AI Systems
The New Uses for SSDs Created by the Latest Inference AI Systems 推論GPUサーバ Inference GPU Server RAGサーバ RAG Server: Improving Answer Accuracy through Knowledge Retention 1 Resolving DB Generation DB Search system memory constraints DRAM 2 コンテキスト Enabling Large- Context Memory Storage(CMX) Server キャッシュ
Slide 25: KIOXIA CM Series – High Bandwidth SSD with TLC Flash Memory
KIOXIA CM Series – High Bandwidth SSD with TLC Flash Memory NVIDIA CMX Server Compatible ◼ Optimized for KV cache workloads ◼ High power efficiency enabled by an in-house controller ◼ Support for the latest PCIe generation and compliance with current standards (NVMe, OCP) ◼ Compatibility with liquid
Slide 26: KIOXIA GP Series – High Performance SSD with XL-Flash
KIOXIA GP Series – High Performance SSD with XL-Flash Compatible with NVIDIA Storage‐Next Super High IOPS SSD ◼ Breakthrough performance from a dedicated SoC combined with XL- FLASH ◼ 100M+ IOPS, ultra-fast and low- latency performance to accelerate Next-Gen AI ◼ Flash memory complements HBM for hig
Slide 27: KIOXIA LC Series – High Capacity SSD with QLC Flash Memory
KIOXIA LC Series – High Capacity SSD with QLC Flash Memory Achieving Next Gen AI Platform ◼ Industry-leading capacity using BiCS FLASH generation 8, 2Tb QLC ◼ 32-die stacking in a compact NAND package through advanced design and integration ◼ 245TB ultra-high capacity in a standard E3.L form factor,
Slide 28: KIOXIA's SSD Lineup Supporting the Growth of Inference AI Systems
KIOXIA's SSD Lineup Supporting the Growth of Inference AI Systems Flash Memory Market For Data Centers (EB) KIOXIA CM Series Inference AI System High Bandwidth SSD Training AI System 909 Traditional Enterprise System with TLC Flash Memory Inference 86% KIOXIA GP Series CAGR High Performance SSD with
Slide 29: Advancing Technology Leadership with BiCS FLASH generation 10
Advancing Technology Leadership with BiCS FLASH generation 10 Atsushi Inoue Executive Officer (Vice President of Memory Division)
Slide 30: Agenda
Agenda ◼ The Evolution of AI Inference Systems and the Dual-Axis Strategy for Response ◼ Technological Leadership of BiCS FLASH generation 10 ◼ Development Status of BiCS FLASH generation 10 © 2026 KIOXIA Holdings Corporation. All Rights Reserved. 30
Slide 31: Expansion of AI Inference Systems and Demands on SSD/NAND
Expansion of AI Inference Systems and Demands on SSD/NAND As AI inference systems evolve and the requirements for SSDs diversify, Kioxia continues to develop low-power memory products optimized for individual applications 推論GPUサーバ Inference GPU Server RAGサーバ RAG Server Low Power DB Generation DB Sea
Slide 32: BiCS FLASH Road Map
BiCS FLASH Road Map Kioxia will further advance its dual-axis strategy by leveraging CBA technology Through optimized Capex, Kioxia will maximize GB output to meet diverse market demands Utilizing CBA Gen.Y Gen.11 High performance by combining optimal cell technology and optimal CMOS circuit 4.8Gbps
Slide 33: BiCS FLASH generation 10: Technological Leadership (Layer Count)
BiCS FLASH generation 10: Technological Leadership (Layer Count) Kioxia achieves cost competitiveness and improved power consumption by optimizing the layer count to balance chip size, investment costs, and power efficiency GB cost vs. Layer Count Advantages of Choosing 332 Layers in BiCS FLASH gene
Slide 34: BiCS FLASH generation 10: Technological Leadership (CBA)
BiCS FLASH generation 10: Technological Leadership (CBA) By early adoption of CBA technology, Kioxia will be able to supply high-performance products superior to those rest of the industry in the future Over 4 years ahead of the industry standard in CBA adoption CY ‘23 ‘24 ‘25 ‘26 ‘27 ‘28 ‘29 CBA ≦3
Slide 35: BiCS FLASH generation 10: Development Status
BiCS FLASH generation 10: Development Status Development of the BiCS FLASH is on track BiCS FLASH Sample Shipments generation 10 Targeted for Summer 2026 1Tb TLC Bit Density +59% Interface Speed +33% Read Throughput >15% better Write Throughput >30% better Read Power Efficiency >40% better High Band
Slide 36: Financial Strategy to Support Our Growth
Financial Strategy to Support Our Growth Yoshihiko Kawamura, Chief Financial Officer
Slide 37: Agenda
Agenda ◼ Elevating Our Quality of Profit ◼ Structural Business Transformation ◼ Achieving Best-in-Class Capital Efficiency ◼ Maximizing Cash Generation and Strengthening Financial Position ◼ Capital Allocation and Shareholder Return Frameworks © 2026 KIOXIA Holdings Corporation. All Rights Reserved.
Slide 38: Elevating Our Quality of Profit
Elevating Our Quality of Profit Capturing demand in the data center market to build a structurally stable revenue model Enhancing the quality of earnings through Capex, R&D and Human capital investments in high-growth, high-margin sectors Market Technology Leadership Cost Competitiveness ✓ Strategic
Slide 39: Structural Business Transformation: Shifting Stable, High-Margin Growth
Structural Business Transformation: Shifting Stable, High-Margin Growth Structurally higher revenue and margin trajectory Drivers for Transformation 100,000 Reducing earnings volatility 90,000 Revenue 90.0% From consumer-driven cyclicality Non-GAAP Operating Margin Market to AI infrastructure-led st
Slide 40: Achieving Best-in-Class Capital Efficiency
Achieving Best-in-Class Capital Efficiency Improve asset efficiency through strict investment criteria and structurally shift to a high ROIC model ROIC Over 60% Target best-in-class ROIC among manufacturing companies ✓ Concentrate Capex, R&D and Human Capital investments in high-growth, high-margin
Slide 41: Maximizing Cash Generation and Strengthening Financial Position
Maximizing Cash Generation and Strengthening Financial Position Balancing growth investments with financial strength to drive sustainable value creation Assets Liabilities and Equity (B¥) (B¥) Enhancing Asset Efficiency Transformational FY26 Capex: Financial Strengthening 450 B¥ (+60% YoY) FY26Q1: F
Slide 42: Capital Allocation and Shareholder Return Framework
Capital Allocation and Shareholder Return Framework Disciplined Capital Allocation to Enhance Corporate Value Subject to change Excess1 depending on the Cumulative2 FCF business environment. Additional Shareholder shareholder return Return Dividends (Base): Progressive dividend 1 Allocation Priority
Slide 43: KIOXIA
KIOXIA

KIOXIA Holdings — Investor Day (June 2026) on 6ix

Kioxia — Investor Day (June 2026)

Slide 1: Flash Memory Scales AI Inference

Slide 2: Agenda

Slide 3: Disclaimer

Slide 4: Kioxia's Growth Strategy in the AI Inference Era

Slide 5: Agenda

Slide 6: Review of the Past Year

Slide 7: Rapid Growth in Inference Workloads

Slide 8: NVIDIA's Proposed GPU-Connected Storage

Slide 9: Focus on the Data Center & Enterprise Markets

Slide 10: Phase Change

Slide 11: Growth Investments to Enhance Corporate Value

Slide 12: Uplifting the world with "memory"

Slide 13: Flash Memory Market and Resource Allocation

Slide 14: Agenda

Slide 15: Flash Memory Market

Slide 16: Supply-Demand Balance and Revenue Market Forecast

Slide 17: Production Strategy

Slide 18: Production Strategy: Front-End Cost per Gigabyte

Slide 19: R&D Strategy

Slide 20: SSD Drives AI Inference

Slide 21: Agenda

Slide 22: Achievements from FY2025

Slide 23: Challenges Faced by Growing Inference Systems

Slide 24: New Uses for SSDs Created by the Latest Inference AI Systems

Slide 25: KIOXIA CM Series – High Bandwidth SSD with TLC Flash Memory

Slide 26: KIOXIA GP Series – High Performance SSD with XL-Flash

Slide 27: KIOXIA LC Series – High Capacity SSD with QLC Flash Memory

Slide 28: KIOXIA's SSD Lineup Supporting the Growth of Inference AI Systems

Slide 29: Advancing Technology Leadership with BiCS FLASH generation 10

Slide 30: Agenda

Slide 31: Expansion of AI Inference Systems and Demands on SSD/NAND

Slide 32: BiCS FLASH Road Map

Slide 33: BiCS FLASH generation 10: Technological Leadership (Layer Count)

Slide 34: BiCS FLASH generation 10: Technological Leadership (CBA)

Slide 35: BiCS FLASH generation 10: Development Status

Slide 36: Financial Strategy to Support Our Growth

Slide 37: Agenda

Slide 38: Elevating Our Quality of Profit

Slide 39: Structural Business Transformation: Shifting Stable, High-Margin Growth

Slide 40: Achieving Best-in-Class Capital Efficiency

Slide 41: Maximizing Cash Generation and Strengthening Financial Position

Slide 42: Capital Allocation and Shareholder Return Framework

Slide 43: KIOXIA