The SQUALO project aims to improve a hardware and software ecosystem for the efficient execution of artificial intelligence algorithms in the space domain. At the heart of the project is the evolution of the GPU@SAT hardware accelerator, a proprietary IngeniArs’ technology designed to deliver high performance and resilience in harsh environments.

PE_00000013 – CUP: J33C22002830006
Ministerial Decree of 15 March 2022 No. 341 – Managerial Provision 5664/2024, Protocol 292234/2024 dated 27/09/2024

Main Objectives

• Enhance the performance of the GPU@SAT IP core through improvements in memory management and internal scheduling.
• Provide additional debugging information to simplify integration with the software stack.
• Enhance dedicated software tools to facilitate the porting of AI models for execution on the target hardware.
• Integrate the entire system into an FPGA-based development kit (AMD Xilinx ZCU104), including a final demonstrator application.

Expected Results

• A complete, robust and validated ecosystem for space-based artificial intelligence applications.
• A testable and replicable devkit for future use cases.
• A practical demonstrator to support the project’s technological maturity.

Funding

“Funded by the European Union – NextGenerationEU through the Cascade Call of the ALMA MATER STUDIORUM – University of Bologna within the FAIR project funded by the NRRP – Mission 4 Component 2, Investment 1.3: extended partnerships among universities, research centers, and enterprises for research projects funding”, M.D. 341 of 15/03/2022, SQUALO, CUP: J33C22002830006