GSI (GSIT) Investor presentation summary

Innovation and product leadership

• Pioneered high-density, high-performance SRAM memory and developed the Associative Processing Unit (APU) for AI, targeting edge AI, inference, and HPC applications.

• 30 years of collaboration with TSMC and over 100 million SRAM units shipped.

• APU's compute-in-memory architecture delivers massively parallel data processing at ultra-low power, ideal for edge workloads.

• Gemini-I and Gemini-II platforms enable real-time AI for defense, aerospace, and edge applications, with Gemini-II offering significant improvements in speed, memory, and efficiency.

• Newest APU, Plato, under development, will support broad LLM and CNN architectures for edge AI.

Market opportunity and strategy

• Leading provider of high-performance SRAM, including radiation-hardened/tolerant products for aerospace and defense, with ASP up to $30K and gross margins over 90%.

• Edge AI semiconductor market projected to grow at 18% CAGR, from $6.8B in 2025 to $16.4B in 2030.

• Strategy focuses on winning in defense and aerospace edge AI, leveraging established relationships and unique low-power advantages.

• Expanding into larger markets with future APU generations, following edge AI growth trends.

Financial performance and capital position

• FY2025 annual revenue was $20.5M, with 122 employees and 146 patents granted (87 for APU).

• October 2025 capital raise of $47M increased cash and equivalents to $70.7M, with no debt and a market cap of $284.1M as of January 2026.

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

• Operating loss for Q3 FY2026 was $6.9M, offset by $3.6M in interest and other income.