Quanta Cloud Technology (QCT) continues its commitment to delivering cutting-edge solutions for the AI-driven future with ongoing efforts and preparations such as validations to support the NVIDIA GB300 NVL72 platform, a next-generation AI infrastructure built on the NVIDIA MGX modular reference architecture.

Production activities for the NVIDIA GB300 NVL72 platform have been well underway in August, with L10 assembled servers and L11 racks undergoing system testing and validation. The first shipment of NVIDIA GB300 NVL72 is also targeted within this timeframe, with volume ramping expected to continue throughout the second half of the year to support global customer deployments.

NVIDIA GB300 NVL72 represents a significant leap forward in AI server design, engineered to meet the escalating demands of AI reasoning, training, and data processing. NVIDIA GB300 NVL72 is a fully liquid-cooled, rack-scale design that unifies 72 NVIDIA Blackwell Ultra GPUs and 36 Arm®-based NVIDIA Grace CPUs in a single platform optimized for test-time scaling inference, poised to become a cornerstone for innovation across industries.

QCT has been a close partner of NVIDIA since the introduction of the NVIDIA Grace Hopper Superchip with the QuantaGrid S74G-2U, NVIDIA GB200 NVL72 with the QuantaGrid D75B-1U, and now the NVIDIA GB300 NVL72 with the QuantaGrid D75U-1U. This long-standing effort has enabled QCT to refine its expertise in liquid cooling, high-power delivery, and system integration with their QuantaGrid servers, which act as the compute trays for next-generation AI infrastructures.

As NVIDIA GB300 NVL72 introduces new architectural efficiencies and unlocks greater AI potential, QCT remains committed to delivering scalable, production-grade QuantaGrid systems that accelerate time-to-market for hyperscalers and enterprise AI adopters alike. By combining NVIDIA’s platform innovation with QCT’s proven engineering and manufacturing capabilities, the two companies are powering the next era of AI infrastructures.