VinAI Research on Friday announced that it now has the most advanced AI supercomputer system NVIDIA® DGX A100 ™, becoming the first owner of a leading-edge AI supercomputer in Southeast Asia.

NVIDIA® DGX A100 ™ is the latest generation of the world’s most advanced AI system released by the AI computing company NVIDIA which can perform five petaFLOPS, equivalent to five quadrillion floating-point operations per second.

Equivalent to a data centre, the supercomputer is designed to provide much more computational performance than the current generation thanks to the NVIDIA Ampere GPU architecture which is built upon the new 7nm process node technology.

VinAI Institute will use this supercomputer for experiments with outstanding large AI models for linguistic data, images and videos for intensive research.

With superior performance, supercomputers will optimise VinAI's computing power and resources by many times with greater accuracy, as well as reduce time required to complete an experiment from more than one week to less than 24 hours.

“With the addition of the new NVIDIA DGX A100 AI supercomputer to our AI computing infrastructure system, VinAI is ready to expand the scale of research and accelerate the deployment of AI application projects,” said Dr. Bui Hai Hung, director of VinAI Research.

He said about 70 researchers, AI trainees and AI engineers will make the most of supercomputers for their current projects.

Dennis Ang, NVIDIA Vice President of Sales and Marketing for Asia Pacific, said: “The new NVIDIA® DGX A100 ™ model will help VinAI optimise its power and computing resources to accelerate diverse workloads, including data analytics, training and inference. VinAI can expand easily as the demand for AI model training increases.”

“We will work closely with VinAI on an architectural level and coordinate with VinAI from the perspective of a product developer in order to promote outstanding results for their AI projects,” he added.

Established in April 2019 with funding from Vingroup, VinAI Research is focusing on basic research in machine learning, deep learning and new methods and applications of AI in computer vision and natural language processing.

Furthermore, the research institute also aims to provide intelligent solutions using AI to solve practical problems in addition to orientation and training of AI talents in the region. — VNS