UTM launches GPU supercomputer for AI research
Universiti Teknologi Malaysia launched high-performance computing infrastructure built on GPUs on 2 October, at a Vice-Chancellor's Gathering at UTM Kuala Lumpur. According to Bernama, UTM is the first public university in Malaysia with such a system, described as the fastest GPU supercomputer among public universities.
Key points
- The system uses NVIDIA Blackwell architecture and can train and run AI models with up to one trillion parameters.
- Processing capacity reaches up to 72 PFLOPS for training and 144 PFLOPS for inference, over a 400 Gbps InfiniBand network.
- An AI workstation laboratory with 120 desktop computers lets up to 120 students and researchers train models at once.
- It is part of a RM20 million Ministry of Higher Education investment under Budget 2024 that supported the Faculty of Artificial Intelligence.
Why it matters for Malaysia
Access to serious computing power is one of the practical limits on university AI research, so a facility on campus gives Malaysian students and researchers hands-on experience with industry-grade hardware.
Questions answered
- What did UTM launch?
- Universiti Teknologi Malaysia launched GPU-based high-performance computing infrastructure at its Faculty of Artificial Intelligence on 2 October 2026. It can train and run AI models with up to one trillion parameters.
- What will UTM's GPU supercomputer be used for?
- UTM's vice-chancellor, Prof Dr Mohd Shafry Mohd Rahim, said it will support AI research, learning and innovation. Planned applications include large language models, computer vision, healthcare, smart manufacturing and systems based on local languages.