GSI Technology reported net revenues of $6.1 million for the fiscal third quarter ended December 31, 2025, up from $5.4 million in the year-ago quarter, representing 12% growth. On a fiscal year-to-date basis, revenues are up 28.5%. Gross margin was 52.7%, down from 54.0% a year earlier, primarily due to product mix shifts. Operating expenses rose to $10.1 million, driven by R&D investments of $7.5 million related to the Plato program and associated IP purchases and consulting fees. The company recorded an operating loss of $6.9 million and a net loss of $3.0 million, or $0.09 per share, compared with a net loss of $4.0 million, or $0.16 per share, in the prior-year quarter. For the fiscal fourth quarter, management guided to revenues of $5.7 million to $6.5 million and gross margin of approximately 54% to 56%.

As of December 31, 2025, GSI held $70.7 million in cash and equivalents, up from $25.3 million at the end of Q2 and $13.4 million at the end of fiscal Q4 2025. Net cash used in operating activities was $7.9 million, with $296,000 used in investing and $53.5 million provided by financing. The cash increase primarily reflects $46.9 million in net proceeds from the October 22, 2025 registered direct offering. Working capital expanded to $71.7 million from $16.4 million at March 31, 2025, and stockholders’ equity rose to $83.6 million from $28.2 million at fiscal year end.

GSI’s Gemini-II APU achieved a three-second time-to-first-token (TTFT) benchmark on a multimodal large language model with text and video input at approximately 30 watts of system power. Third-party testing showed up to three times faster first-token performance versus competitive platforms at lower power. Management emphasized that low power consumption and low latency are critical for edge AI applications, such as drone-based video surveillance, where a TTFT threshold of three seconds enables timely event capture. The company plans further TTFT improvements over the next five months in advance of a Sentinel demonstration.

GSI finalized a proof-of-concept agreement with Israel-based AI firm G2 Tech for an autonomous perimeter security program called Sentinel, involving drones and real-time video analytics. The company expects to receive over $1 million in government funding, which will offset R&D expenses. These funds will support software development for Gemma 3 12B on Gemini-II ahead of a demonstration to government agencies later in the year. G2 Tech selected Gemini-II based on its lowest-power, lowest-latency performance, and a successful POC could lead to design wins in defense and unmanned systems markets beyond the current sponsors.