JPMorgan Announces Major NVIDIA Update: All 6 Next-Gen Vera Rubin AI Chips Enter Final Pre-Production at TSMC Amid Surging AI Demand

JPMorgan Announces Major NVIDIA Update: All 6 Next-Gen Vera Rubin AI Chips Enter Final Pre-Production at TSMC Amid Surging AI Demand

This article does not constitute investment advice. The author holds no positions in the stocks discussed.

JPMorgan Analyst’s Positive Outlook on NVIDIA

Harlan Sur, an analyst at JPMorgan, recently attended a meeting with Toshiya Hari, NVIDIA’s VP of Investor Relations and Strategic Finance. Following the discussions, Sur has expressed a notably optimistic view regarding NVIDIA’s current-generation GPUs and the expected production timeline for their upcoming Vera Rubin platform.

Lead Times Indicate Strong Demand

Sur highlighted that the lead times for NVIDIA’s current-generation Blackwell Ultra GPUs are established in a pattern of “quarters, not months.”This situation is particularly impressive considering the significant ramp-up experienced by Blackwell Ultra in NVIDIA’s fiscal Q2, which now makes up roughly 50% of the overall Blackwell product lineup.

The analyst noted that these prolonged lead times, well over two years into the current AI investment cycle, indicate a scenario where “demand continues to exceed supply.”

Vera Rubin Production Update

In a significant update from NVIDIA, all six Vera Rubin GPUs are reportedly in the final stages of pre-production. This contradicts earlier rumors suggesting delays. According to NVIDIA’s confirmation, the Vera Rubin platform remains on schedule for a planned launch in the second half of 2026.

“[NVIDIA] confirmed that its upcoming Vera Rubin platform has not experienced any delays (despite recent noise to the contrary) and is on track for a C2H26 launch, with all 6 chips that make up the platform having already taped out at TSMC.”

Emerging Demand from Chinese Tech Giants

Recent reports from Reuters indicate growing enthusiasm among major Chinese tech companies, including ByteDance and Alibaba, regarding NVIDIA’s forthcoming Blackwell-based GPU tailored for the Chinese market. These companies have shown a willingness to pay up to double the price of existing H20 GPUs for the projected sixfold performance increase delivered by the new B30A model.

This enthusiasm is reflected in aggregated research from U. S.brokerage firms, which shows a significant preference for NVIDIA’s offerings in China, largely attributed to the superior software support, particularly via the CUDA ecosystem. Additionally, NVIDIA’s GPUs demonstrate enhanced performance in clustered environments, thanks to the innovative NVLink interconnect technology.

Understanding NVIDIA’s Product Range in China

It’s important to mention that the NVIDIA RTX Pro 6000D systems, based on the B40 architecture, can be sold in China without the need for a separate license. This exemption applies since these systems do not incorporate high-bandwidth memory (HBM) and are primarily designed for inference rather than the training of foundational AI models. Given this context, it is highly likely that these chips will see substantial demand once they hit the Chinese market.

For those unfamiliar with the terminology, the B30 GPU is a streamlined version compared to the higher-end B40 GPU within the Blackwell lineup. The B30 is designed to excel in scaling and multi-chip cluster environments, employing dynamic compression to mitigate lower per-chip performance, while the B40 aims to fill the gap left by the banned H20 chip.

Source & Images

Leave a Reply

Your email address will not be published. Required fields are marked *