Luisa Crawford
Apr 12, 2025 00:36
NVIDIA collaborates with SoftBank to rapidly deploy AI factories using DGX SuperPOD technology, marking a significant step in Japan’s AI innovation landscape.
In a notable advancement in AI technology, NVIDIA, in collaboration with SoftBank, is accelerating the deployment of AI factories in Japan using the cutting-edge NVIDIA DGX SuperPOD. This initiative is part of SoftBank Corporation’s ambitious vision to integrate AI into societal infrastructure, according to NVIDIA.
Rapid Deployment and Collaborative Efforts
The AI factory, located in Japan, has been developed at an unprecedented pace. This facility, which did not exist a year ago, now stands as a testament to SoftBank’s commitment to AI-driven innovation. The collaboration with NVIDIA has resulted in the creation of two of the most advanced computing clusters globally. These clusters are designed to process massive data volumes at high speeds, crucial for developing large language models by SoftBank subsidiary SB Intuitions.
NVIDIA’s Infrastructure Specialists (NVIS) played a pivotal role in this achievement. Utilizing their expertise, NVIS ensured that the AI infrastructure was production-ready in record time, transforming bare metal into a functional AI system swiftly and efficiently.
Precision and Agility in Overcoming Challenges
The deployment project faced several challenges, including limited power availability and connectivity issues. However, the NVIS team managed these obstacles with precision, conducting tests during off-hours and creatively repurposing parts from one cluster to maintain the schedule of another. This meticulous management ensured the deployment remained on track.
The DGX SuperPOD deployment involved connecting 510 NVIDIA DGX B200 systems, resulting in impressive processing capabilities. This infrastructure marks a new standard for AI development in Japan, positioning SoftBank as a leader in the AI ecosystem.
Impact on Japan’s AI Ecosystem
SoftBank’s AI factory is not merely a technical feat; it represents a strategic leap forward for Japan’s AI capabilities. The platform is set to support both internal projects like LLM development and provide resources to external companies, thereby enhancing the broader AI ecosystem in Japan.
The deployment exemplifies the NVIS playbook, which focuses on speed, accuracy, and collaboration. By leveraging advanced networking technologies like NVIDIA Quantum-2 InfiniBand and NVIDIA Air, the project ensured smooth operation from start to finish.
Strategic Importance of Speed
For companies adopting NVIDIA’s technology, speed is a strategic asset. Faster deployment reduces downtime costs and accelerates revenue generation from AI-driven products. NVIDIA’s analysis indicates that reducing installation time can save companies significant costs associated with prolonged deployment periods.
This rapid deployment capability allows companies to operate without delays, unlocking revenue opportunities that might otherwise be lost. The collaboration between NVIDIA and SoftBank highlights the potential of AI technologies to drive innovation and efficiency in various industries.
Image source: Shutterstock
