Asia’s lion metropolis is roaring forward in AI.
Singtel, a number one communications companies supplier primarily based in Singapore, will convey the NVIDIA AI platform to companies within the island nation and past.
The cell and broadband firm is constructing energy-efficient knowledge facilities throughout Southeast Asia accelerated with NVIDIA Hopper structure GPUs and utilizing NVIDIA AI reference architectures confirmed to ship optimum efficiency.
The info facilities will function sovereign nationwide assets — AI factories that course of the non-public datasets of corporations, startups, universities and governments safely on shore to provide helpful insights.
Singtel’s first AI companies will spin up in Singapore, with future knowledge facilities underneath development in Indonesia and Thailand. From its hub in Singapore, the corporate has operations that stretch from Australia to India.
Trusted Engines of AI
The brand new knowledge facilities will act as trusted engines of generative AI. Probably the most transformative expertise of our time, generative AI and its capability to amplify human intelligence and productiveness are attracting customers worldwide.
Nations are creating massive language fashions tuned to their native dialects, cultures and practices. Singtel sits on the middle of such alternatives amongst Southeast Asia’s vibrant Chinese language, Indian, Malay and different communities.
Singtel’s initiative helps Singapore’s nationwide AI technique to empower its residents with the most recent expertise. The plan requires considerably increasing the nation’s compute infrastructure in addition to its expertise pool of machine studying specialists.
For companies within the area, having a identified, native supplier of those computationally intensive companies supplies a protected, straightforward on-ramp to generative AI. They will improve and personalize their services whereas defending delicate company knowledge.
Taking the Inexperienced Path
Singtel is dedicated to democratizing AI and decarbonizing its operations.
Its newest knowledge facilities are being constructed with an eye fixed to sustainability, together with within the number of supplies and use of liquid cooling. They undertake finest practices to ship lower than 1.3 in PUE, the ability utilization effectiveness metric for knowledge middle effectivity.
Singtel will use its Paragon software program platform to orchestrate how the brand new AI functions work in live performance with its cell and broadband companies. The mixture will allow edge computing companies like powering robots and different autonomous programs from AI fashions working within the cloud.
A Full-Stack Basis
The corporate will supply its clients NVIDIA AI Enterprise, a software program platform for constructing and deploying AI functions, together with generative AI. Singtel may also be an NVIDIA Cloud Associate, delivering optimized AI companies on the NVIDIA platform.
As a result of Singtel’s knowledge facilities use NVIDIA’s confirmed reference architectures for AI computing, customers can make use of its companies, realizing they’re optimized for main AI efficiency.
Singtel already has hands-on expertise delivering edge companies with NVIDIA AI.
Final Could, it demonstrated a digital avatar created with the NVIDIA Omniverse and NVIDIA NeMo platforms that customers might work together with over its 5G community. And in 2021, Singtel delivered GPU companies as a part of a testbed for native authorities companies.
New AI Function for Telcos
Singapore’s service supplier joins pioneers in France, India, Italy and Switzerland deploying AI factories that ship generative AI companies with knowledge sovereignty.
To be taught extra about how Singtel and different telcos are embracing generative AI, register for a session on the subject at NVIDIA GTC. The worldwide AI convention runs March 18-21, beginning with a keynote by NVIDIA founder and CEO Jensen Huang.
