首页 News 正文

The financial reports released so far have revealed a truth: not all chip companies can benefit from the prosperity of artificial intelligence (AI), highlighting the complexity of the semiconductor supply chain and the dominant position of some companies over others in different fields of the industry.
At present, many semiconductor companies have released their second quarter performance reports, with some companies exceeding expectations and others disappointing, indicating the impact of the rise of artificial intelligence on their profits.
The current interest in artificial intelligence revolves around two key terms - Large Language Models (LLM) and Generative Artificial Intelligence (AIGC). LLM requires a large amount of computing resources and data to train, supporting generative artificial intelligence applications such as chatbots from Google and OpenAI.
The tech giants training LLM are still determined to burn money. Meta announced on Wednesday that it expects a "significant" increase in capital expenditures by 2025 to "support our AI research and product development efforts. Microsoft announced this week that its capital expenditures for the second quarter increased by nearly 80% year-on-year to $19 billion.
As tech giants continue to increase their computing resources, these expenditures are a huge driving force for Nvidia, as the company's graphics processing unit (GPU) is the preferred choice for training these LLMs.
Nvidia's competitor AMD has already launched its AI chip MI300X into the market and is starting to see returns. AMD announced on Tuesday that it expects data center GPU revenue to exceed $4.5 billion by 2024, higher than the company's April forecast of $4 billion. The second quarter profit and revenue announced by this chip company exceeded market expectations.
Chip production and semiconductor equipment companies also seem to benefit from the prosperity of artificial intelligence. TSMC, the world's largest semiconductor manufacturer, announced last month that its net profit for the second quarter increased by over 36% year-on-year; The operating profit of South Korean storage chip giant Samsung Electronics in Q2 surged by 1458.2% year-on-year.
Meanwhile, ASML, the Dutch company that produces the lithography machines required for the world's most advanced chips, announced last month that its net pre orders for Q2 increased by 24% year-on-year, highlighting the continued high demand from top global semiconductor manufacturers such as TSMC.
Not riding the wind of AI
It is worth noting that not all semiconductor companies have been boosted by the growth of artificial intelligence investment, as their importance is relatively small in the early stages of AI.
Qualcomm and Arm's stock prices both fell on Wednesday after releasing disappointing performance guidance. Although these two companies have been emphasizing their importance in the application of artificial intelligence, the reality is that their intersection with AI technology is still very limited.
In fact, many companies' chips are based on Arm design blueprints, and most smartphones in the world also use Arm technology semiconductors. However, although many electronic product manufacturers are talking about artificial intelligence smartphones, it has not fundamentally brought higher growth to this chip design company.
Most of Arm's revenue still comes from consumer electronics products, rather than the data center business that helped AMD and Nvidia achieve high-speed growth performance. Analysts suggest that Arm may benefit as more and more devices begin to adopt artificial intelligence technology.
In addition, Qualcomm's chips are mainly used in smartphones such as Samsung, and most of the company's revenue still comes from smartphones. Similar to Arm, Qualcomm's chips are not used in the data centers required for LLM training.
However, the company's chips will be used in Microsoft's upcoming artificial intelligence personal computers. Therefore, artificial intelligence is a long-term game for Qualcomm.
CandyLake.com 系信息发布平台,仅提供信息存储空间服务。
声明:该文观点仅代表作者本人,本文不代表CandyLake.com立场,且不构成建议,请谨慎对待。
您需要登录后才可以回帖 登录 | 立即注册

本版积分规则

lee382 新手上路
  • 粉丝

    0

  • 关注

    0

  • 主题

    0