GSI Technology and partner Space Micro awarded NASA Phase 1 contract for radiation tolerant AI processing in space
(6 April 2021 - GSI Technology) GSI Technology today announced that, along with prime contractor partner Space Micro, it has been awarded a Phase 1 contract by NASA to develop an optimal real-time data sorting inference processing unit (IPU) board for Earth observation missions that require higher processing power.
As sensors quickly advance, they produce large quantities of data that reveal a critical downside for space-based platforms. They require large amounts of mission data and calibration data to be downlinked, and their data products cannot be conveyed in near real-time. High-performance processors, such as the GSI and Space Micro IPU, are resolving these issues.
“We are pleased to have earned this opportunity to begin working with NASA for its Earth observation mission,” said Paul Armijo, Jr., director of aerospace and defense business sector, GSI Technology. “Our highly skilled team of engineers looks forward to collaborating with NASA in support of the advancement of space exploration and the development of high-performance space computing technology.”
The GSI and Space Micro IPU plugs into the main communication bus and offers enhanced parallel processing power, much like a supercomputer cluster. The main system can reconfigure the IPU device to scale up and down, specifying a specific state depending on the needs of the current task. For added performance and redundancy, multiple processors can be included on the same board, with the capability to stack multiple units together.
The cross-cutting radiation tolerant space on-board data processing product resulting from the NASA contract has the potential to enable many potential NASA space missions, including both manned and unmanned landers and experiments.
About GSI Technology
Founded in 1995, GSI Technology, Inc. is a leading provider of SRAM semiconductor memory solutions. The Company recently launched radiation-hardened memory products for extreme environments and the Gemini APU, a memory-centric associative processing unit designed to deliver performance advantages for diverse AI applications. Gemini APU's architecture breaks the Von Neuman model, the foundation of computing since the 1940’s. Featuring parallel data processing with two million-bit processors per chip, Gemini’s massive in-memory processing reduces computation time from minutes to milliseconds, even microseconds. Gemini excels at large (billion item) database search applications, like facial recognition, drug discovery, Elasticsearch, and object detection. Gemini’s scalable format, small footprint and low power consumption, make it an ideal solution for edge applications where rapid, accurate responses are critical.