Slides
Page 1
Lee-Lean Shu, Chairman, President & CEO Doug Schirle, CFO Didier Lasserre, VP of Sales and Investor Relations Q3 Fiscal Year 2026 Cre at or of t he Associat ive Proce ssing Unit (APU) for AI and a le ading provide r of high- performance memory solutions
Page 2
©2026 GSI Technology, Inc. Safe Harbor The statements contained in this presentation that are not purely historical are forward-looking statements within the meaning of Section 21E of the Securities Exchange Act of 1934, as amended, including statements regarding GSI Technology’s expectations, beliefs, intentions, or strategies regarding the future. All forward- looking statements included in this press release are based upon information available to GSI Technology as of the date hereof, and GSI Technology assumes no obligation to update any such forward-looking statements. Forward-looking statements involve a variety of risks and uncertainties, which could cause actual results to differ materially from those projected. These risks include those associated with the normal quarterly and fiscal year-end closing process, as well as the Company’s ongoing strategic review. Examples of risks that could affect our current expectations regarding future revenues and gross margins include those associated with fluctuations in GSI Technology’s operating results; GSI Technology’s historical dependence on sales to a limited number of customers and fluctuations in the mix of customers and products in any period; the rapidly evolving markets for GSI Technology’s products and uncertainty regarding the development of these markets; the need to develop and introduce new products to offset the historical decline in the average unit selling price of GSI Technology’s products; intensive competition; delays or unanticipated costs that may be encountered in the development of new products based on our in-place associative computing technology and the establishment of new markets and customer and partner relationships for the sale of such products; and delays or unexpected challenges related to the establishment of customer relationships and orders for GSI Technology’s radiation-hardened and tolerant semiconductor products. The strategic review is subject to risks related to the process by which GSI Technology evaluates its strategic alternatives, the terms, timing, structure, benefits and costs of any strategic transaction and whether one will be consummated at all and the impact of any strategic transaction on GSI Technology. Many of these risks are currently amplified by and will continue to be amplified by, or in the future may be amplified by, economic and geopolitical conditions, such as rising interest rates, worldwide inflationary pressures, military conflicts, and declines in the global economic environment. Further information regarding these and other risks relating to GSI Technology’s business is contained in the Company’s filings with the Securities and Exchange Commission. 2
Page 3
©2026 GSI Technology, Inc. GSI Technology key highlights 3 1. Includes cash and cash equivalents, short and long-term investments as of December 31, 2025. 2. Based on the closing share price of $7.85 on January 28, 2026, and common stock outstanding of 36,186,203 shares. • Established in 1995 in Silicon Valley; IPO in 2007 • 30 years of collaboration with TSMC • Pioneers in high-density, high-performance SRAM memory • Developers of Associative Processing Unit (APU) AI chip for Edge AI, Inference and HPC • Newest APU chips addressing edge LLM • $175M for APU R&D funded by SRAM business • $47 M capital raise in Oct. 2025 (post Q2 FY’26 report) Legacy of Innovation and Leadership in Memory Solutions 122 146 (87 for APU) $70.7M ( NO DEBT) $20.5M Employees Worldwide Patents Granted Cash and cash equivalents1 FY 2025 Annual Revenue $284.1M Market Cap2 20% Insider Ownership
Page 4
©2026 GSI Technology, Inc. Market Leader in high-performance SRAM 4 Industry leader with largest portfolio of high-performance memory products SigmaQuad and SigmaDDR core business growth drivers SigmaQuad SRAMs recognized for industry-leading density and speeds 3rd and 4th Generation, sole-sourced, fastest off-the-shelf SRAM on market Expanded, High-Margin Revenue Opportunity—Radiation-Hardened/Tolerant SRAMs recognized for high density, low latency, high bandwidth, at low power consumption
Page 5
©2026 GSI Technology, Inc. Radiation-Hardened & Tolerant SRAM 5 for Aerospace, Satellite & Defense Space Semiconductor TAM Radiation Hardened, Radiation Tolerant 2032 $4.8 B 2022 $2.1 B Allied Market Research February 2024 Advantages: • Hardens chips to protect from solar, deep space radiation • Highest density, fastest speed, with “pin-to-pin” replacement • Completed testing and stringent qualification processes Revenue and profitability drivers: • Expand customer base • ASP up to $30K with 90%+ gross margin • Recurring revenue lifecycles of 10+ years; long design cycles
Page 6
©2026 GSI Technology, Inc. APU for edge ai, inference, and HPC 6 Targeting applications Where power and performance matter 2030 $16.4 B 2025 $6.8 B Specialized Edge AI Semiconductor* SAM Sources: Markets and Markets June 2025; StrateView Research, June 2025 APU’s unique Compute-In-Memory (CIM) architecture, comprised of millions of bit processors tightly coupled with high-density SRAM, delivers massively parallel data processing at ultra-low power for edge applications. 18% CAGR *Drone & Satellite AI chips and Enterprise edge data centers
Page 7
©2026 GSI Technology, Inc. APU features and benefits: • CIM eliminates data movement → Dramatic power reduction • Massive parallel bit-level processing → More performance per watt • Precise bit-width control → Maximum efficiency for edge workloads APU: CIM design—data does not travel → No memory bottleneck = Ultra low-power and real-time speed GPU: Compute happens far from memory → Constant data transfers = Higher latency + higher power usage Distributed High-Bandwidth Memory Bit-Level Parallel Compute (in memory) APU—Built for Real-Time, Low-Power AI DRAM (memory) L2 Cache L1 Cache + Registers Compute ComputeCompute 7
Page 8
©2026 GSI Technology, Inc. Gemini-I platform: Pioneering APU Technology 8 • Enhanced SAR applications for planes, military drones, mobile and traditional data centers • Able to process images in UAVs and planes in near real time • Fast Vector Search (FVS) • Object detection • Facial recognition • Molecular search for new drug discovery • Index builds 4x faster than GPU and ~100x faster than CPU • Cryptography Markets & Applications • C and low-level library Software Framework ®
Page 9
©2026 GSI Technology, Inc. Gemini-II: Market & platform 9 • Onboard SAR provides computer vision in military drones, UAVs, and satellites • Allow autonomous drones to navigate in GPS-denied environment • Fastest TTFT with multimodal LLMs for decision-making in edge applications • Extend mission times in Small Weight and Power Systems (SWaP) edge devices Edge Applications & Benefits • Pytorch, Tensorflow, C Software Framework (mass adoption approach) Optimal for the Edge: Compute Where the Data Lives ®
Page 10
©2026 GSI Technology, Inc. Broad LLM Support Compatible with many open- source large language models (LLMs), including state- of-the-art architectures Efficient Quantization Enabling LLMs to work on small, low-power devices—like drones or sensors—by shrinking them down without losing accuracy Seamless CNN Integration Works seamlessly with a broad spectrum of convolutional neural network (CNN) architectures for computer vision Plato’s versatile architecture will empower developers to leverage a wide range of AI models, from state -of-the-art LLMs to custom CNN architectures. Newest APU: Plato Design underway – tape-out expected in 1H 2027 10
Page 11
©2026 GSI Technology, Inc. Factors driving move from data center to the edge • AI workloads moving to the point of collection to reduce bandwidth and response time • Cloud compute costs make continuous remote inference economically inefficient • Military & defense security and mission constraints may prevent data from leaving device • Increasing demand for real-time inference requires local processing where latency matters Edge Market is growing 11 Applications shifting from cloud to edge Edge AI favors architectures that can deliver more compute per watt.
Page 12
©2026 GSI Technology, Inc. APU Market Strategy Start where we can win now – move to larger markets with future APU 12 Gemini-II Military drone and UAVs Plato LLM at the edge Enhanced APUs Follow edge growth Gemini-II Commercialization Strategy Prioritizing • Edge AI deployment in drones and military vehicles Leveraging • Established relationships with defense agencies and contractors • CIM performance advantages in TTFT, GPS denied environments • APU’s architecture unique ultra-low power advantage 2026 2028 Future
Page 13
©2026 GSI Technology, Inc. GSI awarded $3.4 million of SBIR* funds to date • Announced POC with G2 Tech along with DoW and a foreign defense agency valued at ~$1 million • Awarded Space Development Agency Phase II SBIR valued at $715,000 • Completed $1.25 million Space Development Agency Phase II SBIR • Completed $1.1 million Air Force Phase II SBIR • Completed $250K U.S. Army Phase I SBIR Accessing DOW SBIRs & Government funds *SBIR – Small Business Innovation Research is a United States government program to stimulate technological innovation by funding small businesses to engage in federal R&D with the potential for commercialization. The DoW is one of the largest participants in the SBIR program. ** AFWERX is the innovation arm of the US Air Force, leveraging technology from military as well as private industry, academia and other government agencies. Offsetting R&D expenses & Creating APU awareness within DOW Agencies 13 Future DoW opportunities • Phase 1 SBIR with Air Force worth $75K • Next submission for Phase II with U.S. Army valued at $2 million • Additional SBIR proposals in queue for a total pipeline value of $6 - $10 million • Pursuing Broad Agency Announcement (BAA) opportunities with values up to $40 million Pursuing new DoW funding sources – AFWERX STRATFI/TACFI programs for Gemini-II and Plato Strategic funding from partnerships with potential customers
Page 14
©2026 GSI Technology, Inc. YOY revenue growth from SRAM sales to GSI’s largest customers, driven by strong demand for leading AI processors. Q3 FY2026 includes higher R&D reflecting the IP purchase for Plato and associated consulting fees in the quarter. Q3 FY 2026 cash balance increase from ~$47M (net of fees) from an equity offering closed on October 22, 2025. Financial overview 14 improved balance sheet with October 2025 $50M capital raise $13.4 $70.7 Mar. 31, 2025 Dec. 31, 2025 Cash & Cash Equivalents (in millions) $7.0 $6.7 $10.1 Dec. 31, 2024 Sep. 30, 2025 Dec. 31, 2025 Operating Expenses (in millions) $5.4 $6.4 $6.1 Dec. 31, 2024 Sep. 30, 2025 Dec. 31, 2025 Quarterly Net Revenues (in millions)
Page 15
©2026 GSI Technology, Inc. • APU Optimized for Edge AI with Low Latency and Low Power • Focusing on markets where we can win today with our high performance, low power advantage • Demonstrated Gemini-II Traction in Defense & Aerospace • Multiple DoW SBIR contracts and ongoing customer evaluations for high-value applications (drones, satellites, autonomous systems) • “Sentinel” POC to develop an autonomous perimeter security system using drones and cameras • Strong Operating Foundation • Proven volume semiconductor manufacturing model, over 100 M SRAM units shipped to date • 2025 equity raise funds Gemini-II software build-out and initial Plato development 15 Investment Highlights Differentiated CIM Architecture for real-time AI at ultra-low power
Page 16
GSI Technology High Performance Components fo r Le a d in g-Edge Technology GSITechnology.com / IR Contact: GSIT@HaydenIR.com
Page 17
©2026 GSI Technology, Inc. Appendix 17
Page 18
©2026 GSI Technology, Inc. Parameter Gemini-II Gemini-I Spec vs. Gemini-I Spec Process Technology 16 nm 28 nm Operating Frequency 1.4 GHz 2.33X 600 MHz L1 Size (Memory) 768Mb 8X 96Mb L1 <-> BP Data Bandwidth 367Tb/s 1.16X 315Tb/s Die Size 26 mm x 21.5 mm 2X 20 mm x 14 mm Package Size 37.5 mm x 37.5 mm 2.25X 25 mm x 25 mm 18 Gemini-iI VERSUS Gemini-i
Page 19
©2026 GSI Technology, Inc. APU Platform & roadmap 19 Gemini-I (28 nm) Frameworks Target Applications Libraries APU Roadmap 2015–2026 CY 2026 Securing FundingChip Launch Timing Plato (12 nm)Gemini-II (16 nm)Gemini-I (28 nm) DB Index Builds SAR Fast Vector Search Edge Search SAR “On-board” Computer Vision Multimodal LLMs HPC/Data Search Edge AI, including: Multimodal GenAI/LLMs Computer Vision for • Drones • Automotive • Surveillance • Robotics Search Library (GSL) Vector Math • Radar Library (GRL) • Vector Math • Tensor Math TensorFlowPythonC
Page 20
©2026 GSI Technology, Inc. Gemini-ii: Ultra-Efficient Edge AI Delivers Gpu-class capability at fraction of the power Feature GSI Gemini-II (APU) NVIDIA Jetson AGX Orin Architecture Compute-in-Memory (Associative Processing Unit) GPU + CPU + DLA SoC (Ampere Architecture) Target Segment Low-power, real-time Edge AI (Drones, Defense, Satellites) Embedded AI / Robotics / Industrial Systems Typical Power Use ≈ 15 W (Fixed, Battery-Friendly) 15–60 W Configurable (Maximum performance mode ≈ 60 W) Performance Focus Memory-bound AI / Retrieval, 10× Gemini-I Speed @ 1/10 Power Throughput-heavy DNN Inference, up to 275 TOPS Peak Ideal Use Cases UAVs, Unmanned Vehicles, Tactical Edge Systems Robotics, Industrial AI, Smart Machines • GPU-class capability at a fraction of the power • Optimized for real-time, low-latency edge AI where energy and size are critical • ~15W vs. 60W highlights a 4× power efficiency advantage 20
Page 21
©2026 GSI Technology, Inc. Experienced Management Team Name Title Years of Experience Years with GSI Prior Companies Lee-Lean Shu Chairman and CEO 47 30 Sony, AMD Doug Schirle Chief Financial Officer 47 26 Cypress, Pericom Didier Lasserre VP Sales and IR 38 28 Cypress, Solectron Avidan Akerib VP of Associative Computing 45 10 MikaMonu, NeoMagic Patrick Chaung SR VP of Memory Design 51 16 Sony, AMD Bor-Tay Wu VP of Taiwan Operations 45 29 Atalent, AMD 21
Page 22
©2026 GSI Technology, Inc. Income Statement CONDENSED CONSOLIDATED STATEMENTS OF OPERATIONS (in thousands, except per share data) (Unaudited) 22 Three Months Ended Nine Months Ended Dec. 31, Sept. 30, Dec. 31, Dec. 31, Dec. 31, 2025 2025 2024 2025 2024 Net revenues $6,076 $6,444 $5,414 $18,803 $14,635 Cost of goods sold 2,876 2,911 2,491 8,419 7,794 Gross profit 3,200 3,533 2,923 10,384 6,841 Operating expenses: Research & development 7,457 3,768 4,037 14,322 13,039 Selling, general and administrative 2,649 2,952 2,997 8,331 8,154 Gain from sale of assets - - (56) - (5,793) Total operating expenses 10,106 6,720 6,978 22,653 15,400 Operating loss (6,906) (3,187) (4,055) (12,269) (8,559) Interest and other income, net 3,635 43 70 3,691 274 Loss before income taxes (3,271) (3,144) (3,985) (8,578) (8,285) Provision (benefit) for income taxes (251) 41 44 (156) 124 Net loss ($3,020) ($3,185) ($4,029) ($8,422) ($8,409) Net loss per share, basic ($0.09) ($0.11) ($0.16) ($0.28) ($0.33) Net loss per share, diluted ($0.09) ($0.11) ($0.16) ($0.28) ($0.33)
Page 23
©2026 GSI Technology, Inc. Summary Balance Sheet CONDENSED CONSOLIDATED BALANCE SHEETS (in thousands) (Unaudited) 23 Dec. 31, 2025 March 31, 2025 Cash and cash equivalents $70,672 $13,434 Accounts receivable 2,832 3,169 Inventory 3,876 3,891 Other current assets 1,933 2,961 Net property and equipment 903 808 Operating lease right-of-use assets 8,590 9,547 Other assets 9,654 9,507 Total assets $98,460 $43,317 Current liabilities $7,613 $7,074 Long-term liabilities 7,227 8,017 Stockholders' equity 83,620 28,226 Total liabilities and stockholders' equity $98,460 $43,317