Huang is making money crazily in the field of AI, but AMD also has huge potential: several times higher than imagined

The popularity of ChatGPT has caused the entire IT industry to quickly shift its focus to artificial intelligence. Both overseas and domestic companies are actively developing corresponding products. Companies such as Microsoft and Baidu have launched products with artificial AI. , Google also showed the public the amazing side of its own AI at its recent developer conference. Of course, apart from the major manufacturers, NVIDIA should be the happiest manufacturer. After all, whether it is domestic or overseas companies, almost all AI computing power is provided by NVIDIA’s computing cards. Lao Huang has also made money due to the rise of AI. Get a lot of money.

Everyone is already familiar with the performance of NVIDIA’s computing graphics cards. Whether it is the previous generation A100 or the current H100, they are relatively strong in computing power. Manufacturers also purchase larger quantities at one time to deploy their own AI systems, and NVIDIA The advantage is that the underlying supporting software and algorithms are more suitable for AI. In addition, NVIDIA itself also has a lot of AI hardware and technology in graphics cards, so it is understandable that it is welcomed by users.

Currently in the graphics card industry, NVIDIA and AMD are the two largest companies in GPU, but NVIDIA is indeed close to a monopoly, with an 85% market share in independent graphics. However, AMD is currently the only company that can barely compete with NVIDIA. In addition to game graphics cards, it has also been working hard to launch new products in the computing graphics card sector. AMD previously announced that it would launch the MI300 computing graphics card, but now industry insiders believe that the potential of this computing graphics card may be much greater than we imagined!

The analysis angle of Morgan Stanley analysts is also very strange. At present, no one knows how strong the performance of MI300 is, but Morgan Stanley judges AMD's MI300 based on other information. Morgan Stanley believes that AMD's upcoming MI300 price quotation is relatively high, coupled with recent supply chain evaluations, implying that the potential opportunities are several times higher than previously expected. Morgan Stanley originally conservatively estimated that AMD's AI-related revenue in 2024 would reach US$100 million, and AMD executives were also very conservative and cautious. However, Morgan Stanley now believes that AMD's AI-related revenue will reach US$400 million by 2024. , and may even reach a maximum of US$1.2 billion.

In addition, some of AMD's own recent remarks also show information in the AI ​​field. On the financial call at the beginning of this month, Su Ma admitted that the current sluggishness of the PC market has a negative impact on AMD's development, but she believes that AMD's growth It can get back on track in the second half of the year, and artificial intelligence is one of the keys to AMD's future market expansion. Su Ma said that AI is still in its infancy, and the computing performance required will increase significantly, and AMD has good advantages to seize the opportunity. This actually means that AMD will continue to compete with NVIDIA in computing graphics cards, and will at least launch similar products with comparable performance.

However, the entire industry’s current excessive focus on AI seems to have some negative effects. The investment in AI by major IT companies has squeezed out the expenditure on cloud-related servers. At present, major enterprises are accelerating their investment in AI, causing other cloud supply chains to face temporary weak market conditions. You must know that cloud servers involve chips in all aspects, including storage, computing, memory, etc. Since everyone's investment enthusiasm for cloud servers has declined, these chips The purchase volume will also decline, which will eventually cause the semiconductor economy to continue to slump. If the demand for chips in both the PC market and the cloud is not positive, the consequences can be imagined!

The above is the detailed content of Huang is making money crazily in the field of AI, but AMD also has huge potential: several times higher than imagined. For more information, please follow other related articles on the PHP Chinese website!