As the world's most well-known AI chip, Nvidia can be said to have had a good time recently. However, with the changes in the market environment, NVIDIA's AI chip shipments have gradually been affected, and it can be said that the competitive pressure has increased a lot, and it is even facing the "siege" of many chip giants.
With the explosion of AI software such as Chat GPT, the development of artificial intelligence technology has also entered a new stage, which has also led to a rapid increase in the demand for AI chips.
Nvidia is indeed very top in AI chip technology, GPU itself is a key chip in the field of artificial intelligence development, with the accumulation of GPU technology in the past, NVIDIA certainly has some first-mover advantages in the research and development of artificial intelligence chips. It is these that enhance its competitiveness, and it is also one of the reasons why NVIDIA can dominate the artificial intelligence chip market later.
With artificial intelligence chips, Nvidia once made its market value more than $2,000 billion, second only to Microsoft and Apple. But seeing Nvidia soaring all the way to the sky, other industry bigwigs are also a little unable to sit still, such as Intel.
Nvidia launched the B200 in March, which is also the successor to Nvidia's previous H100. After Nvidia launched the H100 in 2022, the popularity skyrocketed and demand exceeded demand for a time. Therefore, the B200 was actually expected to be high, but after the birth of the B200, it had many competitors at the same time.
On April 9, Intel also announced the details of its own artificial intelligence chips. Intel's Gaudi 3 accelerator performs very well, and its speed is 50% faster than that of H100.
Of course, it's not just Intel, but also Google. Google's next-generation accelerator, TPU v5p, has also been launched, and it is reported that it will be dedicated to training generative AI large models. And Microsoft will also launch its first artificial intelligence chip, MAIa 100, this year.
It can be said that in the field of artificial intelligence chips, various giants are gearing up and are bound to get a share of this huge market blue ocean. Then the efforts of these giants in the field of artificial intelligence chips will inevitably affect Nvidia's current share.
In addition to the "comprehensive containment" of Nvidia by U.S. chip companies, the gradual loss of the Chinese market also made Nvidia very uncomfortable, and even caused Nvidia's market value to fall one after another. Under the U.S. export control of advanced AI chips, Nvidia can be said to be "suffering", clearly mastering advanced AI chip technology, and high-end AI chips have few competitors in the global market, but they cannot ship freely.
Although NVIDIA's H20 chip can be shipped, it has caused dissatisfaction among many domestic AI manufacturers, after all, the performance is only 20% of the previous one, but the price has remained at about 80% of the previous one, and the cost performance can be said to have plummeted a lot.
At the same time, Chinese companies have also developed a large number of low-performance AI chips in series to give full play to more efficient AI training capabilities, which also makes the computing power demand of Chinese enterprises for AI chips decline.
In such a situation, it is no wonder that some foreign media will say that the situation of Nvidia has now reversed. However, some people believe that although giants such as Intel and AMD are indeed challenging Nvidia's position, the opportunity to truly surpass Nvidia is still very limited.
Because now many giants are facing a very objective problem, that is, production capacity. Intel's CEO once told the media that it would probably wait until 2027 or 2028 for the artificial intelligence chip production line in Ohio to be put into production.
Regardless of the outcome of the battle between Intel and Nvidia, we as domestic consumers don't care too much, as long as they can provide consumers with high-quality products before discounts. Of course, what we hope most is that domestic AI chips can catch up and achieve the ultimate surpass.
If you agree, please like, welcome to forward, leave a message and share.