This is not investment advice. The author has no position in any of the stocks mentioned. Wccftech.com has a disclosure and ethics policy.
JPMorgan analyst Harlan Sur has come away from an investor group meeting with Toshiya Hari, the VP of IR and strategic finance at NVIDIA, with a fairly upbeat view of the demand profile for NVIDIA's current-gen GPUs and the tentative production cadence for its next-gen Vera Rubin platform.
Sur notes with cheery effervescence the fact that the lead times for NVIDIA's current-gen Blackwell Ultra GPUs remain firmly ensconced in the range of "quarters, not months," and that's despite Blackwell Ultra ramping up sharply in NVIDIA's fiscal Q2, now constituting around 50 percent of the Blackwell mix.
According to the JPMorgan analyst, NVIDIA's stretched lead times more than 2 years into this AI spending cycle suggest that "demand [is] still outstripping supply."
Moreover, in what constitutes a critical tidbit shared by NVIDIA at the meeting, all six Vera Rubin GPUs have entered final pre-production stages, negating reports of a purported delay:
"[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."
Meanwhile, Reuters recently reported of budding excitement within some of China's biggest tech firms, including ByteDance and Alibaba, for NVIDIA's upcoming Blackwell-based, China-specific GPU, with these titans willing to pay as much as double the H20 GPU's price tag to get their hands on the B30A's reported 6x performance boost over its predecessor.
This aligns with aggregate US broker research, which recently found a strong preference for NVIDIA's products in China.
This is largely due to the fact that NVIDIA's GPUs offer superior software support, especially via the CUDA ecosystem. These GPUs also perform better in a cluster, thanks to the company's NVLink interconnect.
Do note that NVIDIA's RTX Pro 6000D systems that are based on the B40 chip do not require a separate license to go on sale in China, as these systems do not use high-bandwidth memory (HBM) and their main use is for inference rather than training foundational AI models. It is, therefore, quite likely that these chips would sell like hot cake once they do become available for Chinese companies.
For the benefit of those who might not be aware, the B30 is a more cut-down version of NVIDIA's Blackwell GPU vs. the B40, which is a relatively higher-end offering.
The B30 focuses on scaling and multi-chip clusters with dynamic compression to compensate for lower performance per chip, while the B40 aims to offer a direct alternative to the banned H20 chip.
Follow Wccftech on Google to get more of our news coverage in your feeds.
