Chinese internet giants are making bold moves to bolster their Artificial Intelligence (AI) endeavors, sparking a massive $5 billion shopping spree for Nvidia’s processing power. With a growing focus on generative AI, companies like Baidu, ByteDance, Tencent, and Alibaba have placed substantial orders for Nvidia’s A800 processors, aiming to secure the hardware needed to power their ambitions.
Rapid orders for Nvidia’s A800 processors
Recent reports reveal that a consortium of Chinese companies has placed collective orders for around 100,000 units of Nvidia’s A800 processor, representing a staggering value of $1 billion. The expected delivery for this order is set for the end of this year. Additionally, these tech giants have committed to placing further bulk orders, amounting to $4 billion, for processors scheduled to be delivered in 2024.
The Nvidia A800 processor, a derivative of the A100 processor used in data centers, emerges as the focal point of these purchases. The United States government’s export restrictions play a pivotal role in this scenario. These restrictions allow Nvidia to exclusively sell processors with chip-to-chip transfer speeds below 600GBps to Chinese companies, effectively eliminating the A100 and A800 processors from the equation.
Stockpiling in the face of demand and restrictions
Amid the uncertainty posed by potential future restrictions and the surging demand for GPUs, Chinese companies are adopting a proactive stance by stockpiling GPUs. By securing these hardware resources, they aim to guarantee their AI endeavors remain on track. Inside sources reveal that ByteDance, for instance, has already amassed over 10,000 Nvidia GPUs and ordered approximately 70,000 A800 chips worth around $700 million for delivery in the upcoming year.
Industry dynamics and Nvidia’s response
While the Chinese tech companies remained tight-lipped on the matter, Nvidia issued a statement addressing the report. The statement highlighted that consumer internet companies and cloud providers consistently invest significant amounts in data center components, often placing orders well in advance. This aligns with the current trend of burgeoning demand for GPUs driven by the rapid advancements in generative AI.
Potential GPU shortages and escalating prices
Industry experts express concerns about a potential recurrence of the 2020-like GPU shortage due to the escalating demand for these resources. As generative AI continues to advance, the demand for GPUs surges alongside it. The fulfillment of this demand is intrinsically tied to Nvidia’s production capacity, creating a potential bottleneck. Notably, the price of the A800 processor has reportedly increased by more than 50% in the hands of distributors.
With the AI landscape rapidly expanding and more companies venturing into this domain, the trend of bulk orders placed with Nvidia is expected to intensify. As companies strive to equip themselves with the necessary hardware for running large language models (LLMs) and other AI applications, the competition for available resources is set to continue.
The convergence of China’s tech giants and Nvidia’s processing power exemplifies the critical role hardware plays in realizing ambitious AI goals. As these companies scramble to secure the required processing capabilities for their AI endeavors, the demand for GPUs continues to surge. The future holds the promise of AI innovations, but the challenge lies in balancing the supply of essential hardware with the ever-increasing appetite for AI-driven capabilities.
