NVIDIA Corporation‘s extraordinary ascent in recent years, propelled by the wave of artificial intelligence (AI), hit a sudden snag last week when the company was entangled in antitrust woes.
However, a glimmer of hope emerged yesterday as news of a potential U.S.-Saudi Arabia deal involving NVIDIA’s chips and optimistic remarks from CEO Jensen Huang regarding Blackwell chip production caused the semiconductor titan’s stock to soar by 8.2% during Wednesday’s trading session.
A Potential Breakthrough: NVIDIA’s Chips in Saudi Arabia
Reports surfaced on Wednesday suggesting that the U.S. government is contemplating granting NVIDIA the approval to export its advanced chips to Saudi Arabia. This move would empower the Middle Eastern nation to utilize sophisticated AI models with NVIDIA’s cutting-edge technology. The anticipated collaboration includes leveraging NVIDIA’s H200 chips, initially integrated into OpenAI’s GPT-4o, a multimodal transformer.
The relaxation of constraints on the distribution of advanced semiconductors would trigger a significant revenue surge for NVIDIA, driven by heightened demand from Saudi Arabia. NVIDIA is capitalizing on the uptrend in global semiconductor sales, which surged to $51.3 billion in July, marking an 18.7% year-over-year increase, as reported by the Semiconductor Industry Association.
Huang’s Upbeat Outlook on Blackwell
During a technology conference hosted by The Goldman Sachs Group, Inc., Huang revealed that NVIDIA has expedited the production of the eagerly anticipated next-gen Blackwell chips. The company is set to commence shipping these high-performance chips in significant quantities in the fourth quarter to meet escalating demand.
Alphabet Inc., Microsoft Corporation, and Meta Platforms, Inc. have already placed substantial orders for Blackwell chips worth billions. In a similar vein, Amazon.com, Inc. is contemplating a shift from Hopper to Blackwell owing to the latter’s superior AI throughput capabilities.
The latest iteration of Blackwell boasts the NVLink for high-speed communication and a more potent graphic processing unit (GPU). Furthermore, Huang affirmed plans to explore alternative fabrications if necessitated, thereby reducing reliance on Taiwan Semiconductor Manufacturing Company Limited for chip design development.
The NVIDIA Edge in GPU Space
Huang’s recent statement alludes to a substantial data migration from central processing units to GPUs. This is especially beneficial for NVIDIA as the chip giant commands over 80% of the GPU market share, a segment poised to swell from $75.77 billion in 2024 to $1,414.39 billion in 2034, reflecting a 13.8% compound annual growth rate, as per Precedence Research.
NVIDIA’s dominance in the GPU realm has facilitated a significant revenue surge in its data center GPU division, tallying $26.2 billion in the second quarter, a whopping 154% year-over-year uptick. In stark contrast, its arch-competitor, Advanced Micro Devices, Inc., witnessed a relatively modest growth of 115%, accumulating $2.8 billion in revenues from the data center segment.