GSIT News & Events

GSI Technology announced preliminary benchmark results for the Gemini-II Compute-in-Memory processor. These results demonstrated 3-second time-to-first-token performance for multimodal large language models operating at the edge with video and text inputs. "These benchmark results highlight what compute-in-memory can enable for physical AI," said Lee-Lean Shu, President and Chief Executive Officer of GSI Technology. "Edge deployments require fast response under tight power and thermal limits. A 3-second TTFT means the system can generate an initial response every three seconds, which is generally fast enough to be useful in video-based applications without missing meaningful events. Gemini-II's ability to deliver low-latency multimodal inference at low power supports a broader range of real-time applications, from autonomous systems to intelligent machines operating outside the data center."

GSI Technology announced a new proof-of-concept engagement with two government agencies. GSI is partnering with G2 Tech, an Israel Deep tech AI company, on Sentinel, a program to develop an autonomous perimeter security system that manages drones and cameras in real time for advanced monitoring, detection, and response. The project is jointly backed by the U.S. Department of War, and a foreign government agency. Under the POC, G2 Tech is leading the system-level platform development, focusing on real-time data processing, monitoring, and autonomous response workflows for high-load and mission-critical environments across various unmanned systems. The platform is designed to support autonomous operation while maintaining human oversight and decision-support capabilities. By leveraging the APU's compute-in-memory architecture to process sensor data directly where it resides, Gemini-II provides the Sentinel system with on-device AI inference for real-time responsiveness. Total governmental funding for the POC is on the order of millions of dollars. GSI expects to receive roughly a million dollars to support software optimization and integration of the Gemini-II platform into the Sentinel system.