Nvidia signs third SE Asian telco partnership as it eyes digital growth

The AI chip leader's telco collaborations underscore the potential it sees in the fast-growing southeast Asian digital economy.

Robert Clark, Contributing Editor, Special to Light Reading

April 8, 2024

2 Min Read
Nvidia offices
(Source: Nvidia)

Nvidia has unveiled yet another telco partnership in southeast Asia, with the announcement of a $200 million AI investment in central Java.

In the latest sign that the world's hottest tech firm has made the region a priority, it says it will work with Indonesian telco PT Indosat to build an AI center in the city of Surakarta.

It is Nvidia's third southeast Asian collaboration this year. The ten ASEAN countries, Singapore, Thailand and Indonesia among them, have a combined population of 680 million, GDP of $2.8 trillion, and a digital economy that is forecast to be worth $1 trillion by 2030.

Announcing the plan last week, Information and Communication Technology Minister Budi Arie Setiad said the two companies were committed to investing in Indonesian infrastructure and human resources.

Construction of the AI center is part of Nvidia's five-year cooperation agreement with Indonesia, the statement said.

Indosat, Indonesia's number two mobile operator, said last month it would integrate Nvidia's new Blackwell chip into its infrastructure.

AI supercomputer

Nvidia has also supplied its technology to Malaysian conglomerate YTL, enabling it to offer AI cloud services on one of the world's most powerful supercomputers.

The YTL group, which has $15 billion in property, power and construction assets, including mobile operator YTL Communications, has formed a new subsidiary, YTL AI Cloud, which will provide massive-scale GPU-based accelerated computing.

The YTL AI supercomputer will deliver more than 300 exaflops of AI compute, making it one of the fastest supercomputers available. 

Nvidia CEO Jensen Huang said it was one of the first deployments of the Grace Blackwell Superchip on DGX Cloud.

Nvidia's most active commercial partnership in the region is with Singtel, which will offer GPU-as-a-service starting in the third quarter.

Singtel said that from early next year it would deploy on the new Grace Blackwell chips, which deliver 30 times faster real-time large language model inference than its predecessor.

Singtel's Nxera data center unit will build three new sustainable AI-ready data centers in Singapore, Thailand and Indonesia with more than 100MW capacity to support the service.

Bill Chang, CEO of Singtel's Digital InfraCo unit and Nxera regional data center business, said it would allow customers options of different types of accelerators for their advanced computing and AI needs.

He said the company was seeing keen interest from enterprise and government customers "raring to deploy AI at scale quickly and cost-effectively."

Read more about:


About the Author(s)

Robert Clark

Contributing Editor, Special to Light Reading

Robert Clark is an independent technology editor and researcher based in Hong Kong. In addition to contributing to Light Reading, he also has his own blog,  Electric Speech (http://www.electricspeech.com). 

Subscribe and receive the latest news from the industry.
Join 62,000+ members. Yes it's completely free.

You May Also Like