Abstract

This paper presents a proposal for improving the data processing workflow within an AI-powered supercomputing framework for the Atacama Large Millimetre/submillimeter Array (ALMA) Observatory. This framework aims to reduce processing times by 40% to 60%, improve image resolution, and increase data calibration accuracy. A high-performance computing (HPC) architecture with a GPU and artificial intelligence (AI) hardware for astronomical data management is described. The importance of such infrastructure for scientific advancement in developing countries is highlighted by analyzing implementation strategies, cost-benefit analyses, and technology transfer.