(From left) Jongse Park, professor within the School of Computing at KAIST; grasp’s college students Jaehong Cho and Hyunmin Choi; and Brandon Reagen, professor at New York University. Credit: KAIST
■ KAIST introduced on the twenty ninth {that a} research on a big language mannequin (LLM) service infrastructure simulator developed by a analysis group led by Professor Jongse Park within the School of Computing obtained the Best Paper Award at ISPASS 2026, a convention within the subject of laptop methods performance evaluation. Before constructing an precise large-scale synthetic intelligence (AI) server, the group developed a virtual testbed, “LLMServingSim 2.0,” that may verify performance and effectivity in a simulated surroundings. The simulator is anticipated to be broadly used not solely by educational researchers but additionally by LLM service suppliers and AI semiconductor startups to design and optimize next-generation AI infrastructure.
■ Daegu Gyeongbuk Institute of Science and Technology (DGIST) introduced on the twenty ninth that its Next-Generation Sensor and Semiconductor Research Institute has been chosen for a brand new undertaking beneath the “2026 Nano and Materials Technology Development Program – Infrastructure Construction (Fab Advancement)” and the “High-Experience Professional Workforce Utilization Project for Supporting Semiconductor Technology Development,” organized by the Ministry of Science and ICT and the National Research Foundation of Korea. DGIST plans to take the lead in strengthening trade–academia–analysis technical assist primarily based on public semiconductor fabs, together with semiconductor course of expertise growth and fab operation effectivity enchancment; stabilization of kit and fab facility operation; assist for gadget, design, evaluation, characterization, and MPW (multi-project wafer) fabrication; and consulting and on-site technical assist associated to semiconductor expertise growth.
Copyright ⓒ DongA Science. All rights reserved.