At the regular press conference on February 27, a reporter asked a question that according to reports, Nvidia recently submitted to the U.S. ** Trading Commission, for the first time listed Huawei as a major competitor in many categories such as AI chips, and said that if the United States increases restrictions on chip exports, it will further damage Nvidia's competitiveness. Do you have any comment?
In this regard, spokesman Mao Ning saidFacts have proved that the "small courtyard and high wall" cannot stop the pace of China's innovation and development, nor is it conducive to the healthy development of the entire industry, including American enterprises. Openness and cooperation are the core driving force of the semiconductor industry, and China is one of the world's major semiconductor markets. The US should abide by the principles of market economy and fair competition, and support enterprises of other countries in promoting the development and progress of science and technology through healthy competition.
Nvidia listed Huawei as a competitor
Nvidia ranks Huawei as a top competitor in several areas, including a key production area for processors that power artificial intelligence (AI) systems.
The Santa Clara-based company said in its annual report on Wednesday that Huawei is a competitor in four of the five main categories of its business: hardware and software for graphics processing units (GPUs), cloud services, Arm-based processing units (CPUs) and networking products.
Other companies that are also listed as competitors in some areas include AMD, Amazon, Microsoft, and Broadcom.
Just two months before the name Huawei, Nvidia (NVDA) CEO Jensen Huang told reporters in Singapore that the Chinese tech giant was a "strong" competitor to produce artificial intelligence chips, Reuters reported.
However, as the geopolitical relationship between China and the United States has changed, Nvidia has expressed concern about further tensions.
In its annual report, the company said that its competitive position has been damaged if the U.S. export controls on chips change further and could be further affected in the long run.
If [U.S. export control rules] undergo this change, we may not be able to stock such products and may not be able to develop replacement products that are not subject to licensing requirements, which effectively excludes us from all or part of the Chinese market as well as other affected markets, including the Middle East," Nvidia said.
Nvidia reported strong earnings on Wednesday. Profit for the three months ended Jan. 28 was up 769% from a year earlier. But its China business has been hit by U.S. restrictions on the sale of chips to China.
"Growth was strong in all regions except China, where revenue from our data centers dropped significantly following the implementation of export control regulations in the United States** in October 2022," NVIDIA CFO Colette Kress said during the earnings call. ”
Data centers, including graphics cards that are widely used in industrial intelligence, are NVIDIA's largest revenue. Core data center sales increased 409% year-over-year to a record $18.4 billion in the fourth quarter. She said China's data center revenue accounted for a "mid-single-digit percentage" of the company's data center revenue in the fourth quarter and is expected to remain in a "similar range" in the current quarter.
The chip behind Nvidia's skyrocketing
Normally, one doesn't expect computer components to transform entire businesses and industries, but the graphics processing unit released by NVIDIA in 2023 does just that. The H100 data center chip added more than $1 trillion to Nvidia's value and made the company an AI kingmaker overnight. It shows investors that the boom around industrial intelligence is turning into real revenue, at least for Nvidia and its most important businessmen. The demand for the H100 is so great that some customers have to wait up to six months to receive it.
What is NVIDIA's H100 chip?
The H100 is a graphics processor whose name is a tribute to computer science pioneer Grace Hopper. It is an enhanced version of one of the chips that are usually installed in PCs to help gamers get the most realistic visual experience. But it's optimized to process large amounts of data and computation at high speeds, making it ideal for power-hungry tasks of training AI models. Founded in 1993, NVIDIA pioneered this market after nearly two decades of investment, when the company bet that the ability to work in parallel would one day make its chips valuable in applications other than gaming.
Why is the H100 so special?
Generative AI platforms learn to complete tasks such as translating text, summarizing reports, and synthesizing images by training on large amounts of pre-existing material. The more they see, the better they become at things like recognizing human speech or writing cover letters. They develop through trial and error, making billions of attempts to reach proficiency, and consuming a lot of computing power in the process. NVIDIA says the H100 is four times faster at training these so-called large language models (LLMs) than the chip's predecessor, the A100, and responds 30 times faster to user prompts. This performance advantage is crucial for companies racing to train LLMs to perform new assignments.
How did NVIDIA become a leader in artificial intelligence?
The Santa Clara, California-based company is a world leader in graphics chips, which are computer parts that generate the images you see on your screen. The most powerful of these are built with hundreds of processing cores that execute multiple simultaneous computing threads to model complex physics such as shadows and reflections. Engineers at NVIDIA realized in the early 2000s that they could redesign graphics accelerators for other applications by dividing tasks into smaller chunks and then processing them simultaneously. Just over a decade ago, AI researchers discovered that by using this type of chip, their work could finally become practical.
Does NVIDIA have a real competitor?
NVIDIA controls about 80 percent of the market for AI-powered datacenter accelerators operated by Amazon's AWS, Alphabet's Google Cloud, and Microsoft's Azure. These companies work hard to build their own chips in-house, as well as chips from Advanced Micro Devices Incand Intel Corpbut have not yet made much of an impact in the AI accelerator market.
How does NVIDIA get ahead of the competition?
Nvidia quickly updated its products, including hardware-enabled software, at a speed that no other company could match. The company has also designed a variety of clustering systems to help customers purchase H100s in bulk and deploy them quickly. Chips like Intel Xeon processors are capable of more complex data processing, but they have fewer cores and are much slower when processing large amounts of information that are typically used to train AI software. Revenue from NVIDIA's data center division grew 81% to $22 billion in the last quarter of 2023.
How do AMD and Intel compare to NVIDIA?
AMD, the second-largest maker of computer graphics chips, launched a version of its Instinct series in June, targeting a market dominated by NVIDIA products. AMD CEO Lisa Su told the audience at an event in San Francisco that the chip, called the Mi300X, has more memory and can handle workloads for generative AI. "We're still in the very, very early stages of the AI life cycle," she said last December. Intel is bringing to market specific chips for AI workloads, but acknowledges that demand for data center graphics chips is currently growing faster than demand for processor units that have traditionally been its strengths. NVIDIA's strength isn't just in the performance of its hardware. The company invented a language called CUDA, a language used for its graphics chips that allows the graphics chip to be programmed to support the type of work of an AI program.
What's next for nvidia?
Later this year, the H100 will pass the torch to its successor, the H200, after which NVIDIA makes more substantial changes to the design and further introduces the B100 model. CEO Jensen Huang acts as an ambassador for the technology and seeks to get ** and private companies to buy it sooner rather than later.
The Year of Science and Technology in China