The demand for Artificial Intelligence (AI) continues to surge, pushing the boundaries of data center capabilities. Powering these complex AI systems efficiently presents a significant challenge for businesses worldwide. Flex, a leading manufacturing partner, has introduced a groundbreaking solution. This new development significantly enhances Flex AI Infrastructure, ensuring high-performance computing environments operate with unprecedented efficiency.
Revolutionizing Flex AI Infrastructure Power
Flex (NASDAQ: FLEX) recently unveiled a new power shelf system at OCP APAC. This system is specifically designed for the NVIDIA GB300 NVL72 platform. It delivers an impressive 33 kW per shelf. Furthermore, it achieves a remarkable 97.5% peak efficiency at half-load. This innovative design dramatically reduces power conversion losses. In fact, it cuts these losses by over 60% compared to conventional systems. Consequently, data centers can achieve significant energy savings. This also leads to reduced cooling requirements, lowering operational costs.
The new power shelf features six power supply units. These units operate in a 3+3 redundant configuration. This redundancy ensures high reliability and continuous operation, crucial for demanding AI workloads. Moreover, the system is engineered to support NVIDIA’s platform, which incorporates 72 Blackwell Ultra GPUs. Such robust power delivery is vital for supporting the intense computational demands of modern AI models. This advanced power management system truly strengthens Flex AI Infrastructure capabilities.
Strategic Advancements in Flex AI Infrastructure
Flex is actively preparing for the industry’s transition to an 800 VDC data center power infrastructure. This forward-looking approach positions Flex at the forefront of future data center design. This higher voltage infrastructure is essential for enabling future megawatt-scale racks. These racks will become the standard for large-scale AI factories. The system’s readiness aligns perfectly with NVIDIA’s upcoming Kyber rack-scale systems. This collaboration highlights Flex’s commitment to supporting the exponential power demands of advanced AI. Indeed, Flex aims to facilitate seamless scalability for increasingly demanding AI models.
The company plans full-scale production of 800 VDC power components. This production will coincide with the release of NVIDIA Kyber rack-scale systems. Therefore, Flex ensures its solutions are ready for the next generation of AI deployment. This strategic move reinforces Flex’s role as a critical enabler in the evolving AI landscape. Their solutions address the core challenges of power, heat, and scale in the AI era. This proactive development underscores the strength and foresight of Flex AI Infrastructure solutions.
Global Impact and Flex’s Role in AI Growth
The announcement at OCP APAC underscores the global relevance of these advanced power solutions. Asia-Pacific markets are witnessing rapid acceleration in AI infrastructure deployment. Therefore, efficient power management becomes even more critical. By focusing on this vital challenge, Flex establishes itself as a key player in the AI supply chain. They extend beyond just manufacturing, offering comprehensive solutions. Flex provides advanced manufacturing capabilities, innovative power and cooling products, and end-to-end lifecycle services. These services help overcome data center power, heat, and scale challenges worldwide.
Chris Butler, president, Critical and Embedded Power at Flex, emphasized their collaborative efforts. He stated, “Our collaboration with NVIDIA enables hyperscale and colocation customers to deploy next-generation AI infrastructure with greater speed and efficiency.” This partnership is crucial as customers scale toward 1 MW racks and 800 VDC data center power infrastructure. Flex’s robust portfolio of grid-to-chip power products further supports this mission. Ultimately, Flex accelerates data center deployment globally. This commitment solidifies the foundational role of Flex AI Infrastructure in the future of AI.
Frequently Asked Questions (FAQs)
What are the key features of Flex’s new power shelf system for NVIDIA GB300 NVL72?
The system delivers up to 33 kW per shelf. It also features six power supply units in a 3+3 redundant configuration. Furthermore, it achieves 97.5% peak efficiency at half-load, significantly reducing power conversion losses by over 60%.
How efficient is the new Flex power shelf system?
The Flex power shelf system achieves an impressive 97.5% peak efficiency at half-load. This reduces power conversion losses by more than 60% compared to conventional systems. This leads to lower heat output, reduced cooling demands, and decreased energy costs for data centers.
How does Flex support future data center power infrastructure?
Flex is actively preparing for the transition to 800 VDC data center power infrastructure. They plan full-scale production of 800 VDC power components to coincide with the release of NVIDIA Kyber rack-scale systems. This enables future 1 megawatt IT racks for next-generation AI infrastructure.
What are the benefits of Flex’s advanced power management for AI?
Flex’s advanced power management offers numerous benefits. These include industry-leading efficiency, reduced operational costs through lower energy consumption and cooling, high power density, and a robust, redundant power supply. It also ensures readiness for future megawatt-scale AI factories.
When will Flex begin production of 800 VDC power components?
Flex plans to begin full-scale production of 800 VDC power components to coincide with the release of NVIDIA Kyber rack-scale systems. This ensures seamless integration and scalability for advanced AI models.
