GSI (GSIT) Investor presentation summary

Innovation and product leadership

• Developed the Associative Processing Unit (APU) and Compute-in-Memory (CIM) architecture for AI, targeting edge applications with ultra-low power and real-time performance.

• Over $175M invested in APU R&D, with 146 patents granted, 87 related to APU technology.

• Gemini-II APU demonstrates up to 300% better performance-per-watt versus competing edge platforms in multimodal AI inference.

• Plato, the next-generation APU, will support large language models and CNNs for edge AI, with tape-out expected in 1H 2027.

• SRAM business remains a high-margin, industry-leading foundation, funding APU development.

Market opportunity and strategy

• Edge AI market projected to grow from $21B in 2025 to $120B in 2030, with Gemini-II targeting specialized edge AI segments.

• Focus on applications such as drones, satellites, smart city sensors, and enterprise edge inference.

• Commercialization strategy leverages defense sector relationships and proven CIM performance advantages.

• Radiation-hardened SRAM targets aerospace and defense, with recurring revenue and high gross margins.

Financial performance and funding

• LTM revenue as of Q3 FY2026 was $24.7M, with $70.7M in cash and no debt after a $47M capital raise in October 2025.

• Q3 FY2026 net revenues were $6.1M, with increased R&D expenses due to IP purchases for Plato.

• SRAM revenue provides non-dilutive funding for APU development, supporting ongoing innovation.

• Balance sheet strengthened with $83.6M in stockholders' equity as of December 31, 2025.