SBIR/STTR Award attributes
The amount of unstructured data on missions is increasing every minute of every day. This is a huge challenge for the military and commercial companies - especially in the space community - where the data vastly exceeds the capacity limits of the communication systems that transmit the data to the ground. However, this big data challenge can be combated by developing next-gen space hardware specifically designed to efficiently process big data at the edge. The USSF faces incredible challenges with processing big data problem-sets, especially as it pertains to executing the various Space Domain Awareness (SDA) mission-sets with multi-domain assets operating in a contested and dynamic environment. The commander of the U.S. Space Command, Gen. James Dickinson, recently called SDA the command’s “No. 1 need.” He went on to further explain, “We need to enhance our understanding of the congested and complex space operational environment, to include what is occurring and when, and the intent behind those engaged in such actions,” at an 8 March 2022 Senate Armed Services Committee hearing. To make this a reality, the US needs to invest in and develop next gen computing hardware and software specifically designed to operate in harsh space environments. GSI Technology, Inc. (GSI) seeks to solve this problem through our next-generation non Von-Neumann Associative Processing Unit (APU) compute in-memory integrated circuit (IC). Founded in 1995, GSI is a leading provider of semiconductor memory solutions. GSI recently launched radiation-hardened memory products for extreme environments in space and the Gemini® APU1.3, a memory-centric design that delivers significant performance advantages for diverse AI applications. The Gemini APU1.3 architecture established the capability to remove the I/O bottleneck between the processors and memory arrays by performing massive parallel search directly in the memory array where data is stored. The novel architecture delivers performance-over-power ratio improvements for applications like image detection, speech recognition, e-commerce recommendation systems, and more. In this Direct to Phase II effort, GSI will partner with AFRL’s Space Vehicles Directorate - Space Electronics Technologies Branch (RSVS) to fabricate a next-generation Associative Processing Unit-2 (APU2) compute in-memory IC as well as design and fabricate an APU2 Evaluation Board. The overall goal of this development effort is to develope of monolithic, intelligent compute in-memory engines, with an emphasis on processing vast amounts of data in real-time. GSI’s APU2 will be an ideal solution for edge applications with a scalable format, small footprint, and low power consumption where rapid, accurate responses are critical. This will be a critical technology element to enable the USSF to rapidly detect, warn, characterize, attribute, and predict potential and real threats in space; ultimately to maintain and exploit space superiority.
