Photo/VCG

On June 2nd, NVIDIA’s founder and CEO, Jensen Huang, announced that the production of NVIDIA’s Blackwell chips has commenced. During his speech, Huang revealed that NVIDIA will launch the Blackwell Ultra AI chip in 2025. The next-generation AI platform, named Rubin, will utilize HBM4 memory. The Rubin platform is currently under development and is set to be released in 2026, featuring HBM4 memory chips.

The first Blackwell chip, known as GB200, is claimed to be the “most powerful chip in the world” at present. The supply chain has high expectations for the GB200, with projections suggesting that its shipments could exceed one million units by 2025, potentially accounting for nearly 40% to 50% of NVIDIA’s high-end GPU shipments.

Currently, GPU products based on the Blackwell architecture are in production and are expected to be a significant revenue driver for 2024 and 2025. A report by JPMorgan Chase indicates that NVIDIA is expected to maintain growth in the second fiscal quarter, primarily due to continued customer spending on AI/accelerated computing plans and strong demand for its Hopper H100 and new H200 GPU platforms (Blackwell GB200/B200/B100). Blackwell-related new products are anticipated to begin preliminary production in the third fiscal quarter and achieve large-scale shipments in the fourth fiscal quarter.