As Asia’s leading communication technology group, Singtel has been connecting people and businesses for over 140 years. To increase adoption of their 5G enterprise business, Singtel has invested in capabilities such as developing the Singtel Paragon Platform and multi-access edge compute (MEC) services for its 5G network. The MEC infrastructure leverages NVIDIA accelerated compute and NVIDIA AI Enterprise to support AI and graphics-intensive use cases and provide the performance, reliability, and manageability needed by customers to deploy mission-critical applications.
Singtel
Red Hat
Computer Vision / Video Analytics
NVIDIA AI Enterprise
NVIDIA-Certified Systems
Singtel developed their MEC platform so their customers can innovate with business-critical AI applications, including real-time video analytics, AR/VR, and chatbots. These workloads require very low-latency performance and high bandwidth. MEC allows customers to run their compute-intensive AI applications at the edge to provide the necessary experience to the end user.
Singtel’s MEC platform includes NVIDIA T4 Tensor Core and A40 GPUs, and NVIDIA AI Enterprise, the end-to-end software layer of NVIDIA’s AI platform, to provide the performance and reliability needed to enable AI applications across their 5G network.
To support mission-critical applications on the edge, Singtel collaborated with NVIDIA to leverage the power of NVIDIA GPUs to run AI workloads and graphics processing on the MEC. To handle multiple concurrent users, Singtel turned to GPU virtualization with NVIDIA AI Enterprise to effectively share GPU resources across a broad range of AI applications.
With over 100 frameworks, pretrained models, and development tools, NVIDIA AI Enterprise enabled Singtel to quickly develop and test a broad range of AI applications and use cases across their 5G network. As Singtel was evaluating NVIDIA AI Enterprise, they leveraged NVIDIA Riva, a fully customizable, multilingual speech and translation AI SDK, and NeMo to create an AI chatbot demo. As Singtel’s customers move from pilot to production, the collaboration between Singtel and NVIDIA becomes critical because it allows customers to utilize NVIDIA’s AI Enterprise software and AI expertise, and Singtel’s 5G network and Paragon Platform.
With the collaboration and introduction of NVIDIA AI on the MEC, Singtel’s customers can speed up the development, implementation, and manageability of innovative AI applications across 5G. After the successful development of the chatbot demo, Singtel is planning to further leverage the extensive library of frameworks, workflows, and pretrained models to develop additional new and interesting use cases on their 5G platform. For example, customers are already leveraging NVIDIA DeepStream and PyTorch libraries—included in NVIDIA AI Enterprise—for video analytics applications.
As Singtel continues to deploy Singtel Paragon, its orchestration platform globally, these capabilities will be extended to customers outside the region. Because NVIDIA AI Enterprise is certified to deploy anywhere, Singtel’s validation of the solution in its home market will enable regional telcos to leverage this existing success and power of AI and 5G use cases across the globe.
“Having a full software stack and rich ecosystem of tools and libraries is key to supporting the development of 5G applications and use cases. This acceleration with NVIDIA AI Enterprise helps unlock value from Singtel’s 5G network for mission critical use cases.”
Manoj Prasanna Kumar
Head of Technology, Singtel Group Enterprise
Ready to kick-start your enterprise AI journey?
Learn more about NVIDIA AI Enterprise.