Nvidia’s Leading Position

The semiconductor industry is on the rise, with a huge influx of money pouring into the semiconductor ecosystem, from mining, sorting, processing minerals to research and development, and production of chips.

Estimated to be worth nearly trillions of dollars, many companies in this industry from Asia to America, Europe have achieved tremendous success.

For a long time, many “giants” in the semiconductor industry have dominated the top rankings of businesses. Not stopping there, many experts believe that the era of new semiconductor companies is just beginning.

Founded over 30 years ago with its initial focus on computer graphics chips for gaming, Nvidia quickly shifted its focus to AI.

Jensen Huang is one of the most seasoned CEOs in the technology industry. He laid the foundation for Nvidia’s AI development in 2006 when he developed his own chips for purposes beyond computer graphics.

Nvidia is the biggest beneficiary in the race among technology companies to integrate artificial intelligence (AI) into their products and services. Many tech giants are ready to spend billions of dollars to buy Nvidia’s graphics processing units (GPUs).

Bí quyết giúp Nvidia trở thành cỗ máy in tiền, các nước xếp hàng mua chip - 1

CEO Nvidia showcases his logo tattoo on his hand (Photo: Reuters).

Nvidia’s graphics processing units (GPUs), H100, are commonly used to create the most sophisticated AI systems. Nvidia’s H100 is known as the most powerful GPU on the market, and it’s also very expensive, priced at around $30,000-$36,000 USD and can go up to $40,000 USD when scarce (almost 1 billion VND).

However, AI developers have to “wait in line” for months to use Nvidia’s processors through cloud service providers.

Satellite companies in the semiconductor industry, playing a subcontracting role, also own popular companies like TSMC, Samsung, SMIC… with revenues reaching tens of billions of USD per quarter.

Analysts estimate that Nvidia can produce around 1.2 million chips per year, but meeting the demand is becoming increasingly challenging. Nvidia designs chips and contracts with the world’s largest semiconductor company, TSMC, for manufacturing.

However, the company is facing congestion in the subsequent steps of the production process when silicon pieces are assembled into final products. TSMC aims to double its capacity in the next steps this year.

Rapid Growth

Nvidia took 24 years as a public company to reach a valuation of $1 trillion. However, just within 8 months, they achieved another $1 trillion valuation thanks to their role as a chip manufacturer driving the AI revolution. Most recently, Nvidia’s market capitalization has exceeded $2 trillion.

The journey to becoming one of America’s most valuable companies began at a Denny’s store in 1993. The company has seen rapid growth in recent years thanks to Nvidia’s dominance in the GPU arena.

These chips have become rare and precious commodities, with Nvidia currently holding about 80% of the market share. Demand has outstripped supply and prompted competitors to develop their own chip models.

The ability to ensure a supply of GPUs can determine the pace of development of new AI systems for companies. Therefore, many companies are using their stable GPU supply to recruit more AI specialists. These chips are even being used as collateral for borrowing billions of dollars.

Bí quyết giúp Nvidia trở thành cỗ máy in tiền, các nước xếp hàng mua chip - 2

Nvidia’s revenue is growing rapidly (Photo: Nvidia).

Fletcher Previn, CIO of Cisco Networking Technology, says that these chips are so valuable that they are delivered in armored vehicles.

Recently, when Nvidia announced better-than-expected results in the third consecutive quarter, the company’s executives revealed that supply is still tight and the new generation of AI chips expected to be introduced this year will be limited in supply.

Chips that lead to the development of large language models as the foundation for artificial intelligence that can generate language, like OpenAI’s ChatGPT, are being used. Most spending on AI by tech companies like Microsoft, Alphabet, and Amazon.com is dedicated to GPUs.

Jensen Huang, CEO and co-founder of Nvidia, says that generative AI is driving a wave of investment worth trillions of dollars. He believes that this wave will double the number of data centers worldwide in the next 5 years and bring market opportunities for Nvidia.

“A completely new industry is taking shape, and that is driving our growth,” he said at the company’s meeting.

In the financial quarter ending in late January, Nvidia generated $22.1 billion in revenue, up 265% from the same period last year, with net profit nearly 8 times to $12.3 billion. Nvidia’s stock has risen 59% this year after more than tripling in 2023.

Competing to Buy Chips

To cope with the AI competition, four major Chinese tech companies are competing to buy chips from Nvidia. Baidu, ByteDance, Tencent, and Alibaba have purchased about 100,000 A800 chips worth billions of dollars from Nvidia.

Currently, Nvidia’s GPUs, which provide computing power for developing large models, have become essential for major tech companies worldwide.

Due to concerns that the U.S. will tighten chip control and disrupt chip supplies, many companies have no choice but to stockpile A800 chips as no one wants to fall behind in the AI competition.

A Baidu employee said, “Without chips, we won’t be able to train any language models.”

ByteDance, TikTok’s parent company, has many small research groups working on various general-purpose AI products and are undergoing internal testing.

Bí quyết giúp Nvidia trở thành cỗ máy in tiền, các nước xếp hàng mua chip - 3

The number of H100 chips sold and server market revenue expected by 2027 (Photo: Omdia).

Last May, ByteDance tested an AI-powered chatbot called Tako on TikTok that can create videos based on user requests. TikTok recently also filed a trademark application for Tako, and the company may be preparing to promote this chatbot.

ByteDance employees said the company had reserved at least 10,000 Nvidia GPUs to support the development of AI-related products. They also mentioned that the company had ordered nearly 70,000 A800 chips worth around $700 million.

Chinese companies only have a few thousand chips available for training AIs. This strong demand has also led to a shortage of Nvidia chips.

Not only cloud service providers and tech companies are seeking chips, but many AI research labs are also buying chips, and Nvidia is the most ordered chip company.

With increasing demand, chip prices are also rising. A Nvidia distributor revealed, “The price of A800 chips delivered to distributors has increased by over 50%.”

Brian Venturo, co-founder and CTO of CoreWeave, believes this is due to Nvidia’s dual strength in hardware and software that the company has spent many years building for its chips. He believes no company can produce better chips than Nvidia.

Focus on Developing Own Chips

The high demand for chips is also driving Nvidia’s competitors to intensify the development of their own AI chips.

Advanced Micro Devices has begun selling chips to compete with Nvidia’s products, with sales of these chips expected to reach over $3.5 billion this year.

UK chip designer Arm Holdings has also entered the arena. Intel has started selling AI-capable central processing units.

In addition, many AI chip manufacturing startups are also competing in the market. Major cloud computing companies like Google and Amazon are also making efforts to develop AI chips. Last November, Microsoft unveiled its first AI chip, called Maia 100.

Bí quyết giúp Nvidia trở thành cỗ máy in tiền, các nước xếp hàng mua chip - 4

Jensen Huang, CEO and co-founder of Nvidia (Photo: Investing).

Many analysts in the industry believe that Nvidia’s advantage is not easily affected by competition, thanks to the depth and complexity of the software that the company has built for its chips over many years.

Meanwhile, startups as well as major tech companies are all advertising the number of Nvidia chips they have accumulated.

Last month, Mark Zuckerberg, CEO of Meta Platforms, announced on Instagram that his company plans to spend several billion dollars to buy 350,000 Nvidia H100 chips by the end of this year.

Google has set up a steering committee to decide how to allocate computer resources between internal users and outside users. Microsoft has also set up a similar allocation program, known as the GPU Council, where executives determine how to allocate computer resources for internal projects.

However, Andrew Ng, who runs an AI fund, says that AMD and Intel have made significant strides in developing competitive software systems alongside AI support chips.

“I think within about a year, the semiconductor chip shortage situation will improve,” he shared with the Wall Street Journal.

Trên đây là tất cả các thông tin về ” Bí quyết giúp Nvidia trở thành “cỗ máy in tiền”, các nước xếp hàng mua chip” mà Greenpineresort mang đến . Chúc quý độc giả luôn vui vẻ và tràn ngập tình yêu trong cuộc sống. Hãy ghé thăm Greenpineresort một ngày thật sớm bạn nhé!
Booking hoặc đặt lịch vui lòng liên hệ:
SDT: 091 529 93 68
Email :[email protected]