Singtel, a major telecom provider in Singapore, is set to launch a new cloud computing service called GPU-as-a-Service (GPUaaS) with the help of NVIDIA AI chips. The cloud computing service will be released in the third quarter of 2024 and will provide access to graphics processing units (GPUs) on demand.
Singtel's Collaboration with Nvidia AI Chips
According to Straits Times, Singtel's GPUaaS will initially utilize NVIDIA's new AI chip, the H100 Tensor Core GPU clusters, in upgraded data centers in Singapore. It plans to be among the first to deploy NVIDIA's GB200 Grace Blackwell Superchips, which will deliver 30 times faster real-time large language model inference.
"Our GPUaaS will run in AI-ready data centres specifically tailored for intense compute environments with purpose-built liquid-cooling technologies for maximum efficiency and lowest PUE, giving them the flexibility to deploy AI without having to invest and manage expensive data centre infrastructure," stated Bill Chang, CEO of Nxera and Singtel's Digital InfraCo.
This expansion will empower enterprise customers with various accelerator options for their advanced computing and AI requirements.
The service will later extend to new sustainable AI data centers in Singapore, Thailand, and Indonesia, facilitating over 100MW deployment over the next three years through Singtel's regional data center business, Nxera.
In response to Singtel being the first to use the new NVIDIA AI chips, Ronnie Vasishta, Senior Vice President of Telecom at NVIDIA, praised their initiative. He noted that Singtel's GPUaaS, powered by NVIDIA, will enable businesses across Southeast Asia to leverage AI effectively for operational transformation.
Alongside this initiative, Singtel also introduced Paragon-S, a digital aggregation and orchestration platform designed to integrate satellite and terrestrial network connectivity services.
Join the Conversation