Source: Golden Ten Data

Over the past year,$NVIDIA (NVDA.US)$,$Microsoft (MSFT.US)$und$Alphabet-A (GOOGL.US)$Until the US tech giants have benefited from the artificial intelligence boom in the market over the past year. But many small companies in the global AI supply chain will also benefit.

Bank of America said in a note on March 15 that since the second half of 2023, the AI market has been “booming”, and various artificial intelligence solutions have emerged. Bank of America says most chips rely on$Taiwan Semiconductor (TSM.US)$Production, chips, and integrated circuit designs come from around the world or Taiwan, and some components come from Greater China.”

Bank of America Bank analysts wrote, “The scarcity of AI/computing vendors helps increase added value and bargaining power, particularly for expensive artificial intelligence products. The list lists all parts of the supply chain that will benefit from artificial intelligence, as well as popular stocks in every category.

Server original design manufacturer:

The bank notes that global companies have partnered with Taiwanese manufacturers on highly customized projects, such as artificial intelligence GPUs. Here are Bank of America's preferred stocks among such manufacturers or suppliers:

Guangda Computer: The bank believes that Quanta is a top original design manufacturer and has a diverse customer base. Due to its ability to produce highly customized server designs, Quanta has collaborated with Google for over a decade.

Weichuang Capital: According to the bank, Weichuang is a major supplier of CPU substrates. With the increase in AI server applications and GPU supply, the bank expects the AI server GPU substrate business to account for 3% to 6% of Wistron's total sales in 2024 and 2025, compared to only 1% in 2023.

Foxconn Industrial Internet: The bank notes that Foxconn has become an important partner of Microsoft and Nvidia through its subsidiary Ingrasys. It also notes that Foxconn Industrial Internet has supported Microsoft's OpenAI for years.

Hardware technology:

Hardware technology, such as printed circuit boards, is an upstream part of the AI server supply chain. Bank of America said these materials or components are “always needed.” Here are the bank's preferred stocks in this field: These are Bank of America's top picks in this field: Bank of America said that Xinxing Electronics and NYPCB are leading the way in developing technologies and products related to artificial intelligence and high-performance computing. However, according to the bank, Xinxing Electronics is faster in increasing revenue related to AI servers, as can be seen from its “proven record” of producing printed circuit boards for Nvidia.

Semiconductors:

Bank of America said that as chip design becomes more complex, the cost of developing physical components is also increasing, and the cost of chip design is getting higher and higher. However, the bank said that since manufacturing chips for artificial intelligence applications requires more advanced technology, “increasing reliance” on companies with a “reliable record” of designing customized chips is growing. The bank's analyst wrote, “We believe that in the age of artificial intelligence, finding trustworthy intellectual property and design service providers is becoming more important because overall R&D investment costs are getting higher and customers (chip designers) are unwilling to take the risk of failure.” Overall, IP and design services account for only a small portion of the global semiconductor market, but since chip performance is largely dependent on the quality of IP and chip design, their importance is increasing. In this context, the bank listed Taiwanese stocks AlChip and eMemory as the main beneficiaries. Additionally, MediaTek (MediaTek) is a leading waferless company in Asia and an indirect beneficiary of the AI server trend. The bank said MediaTek has “strong room for growth” in expanding the application of its custom chips in hyperscale data centers. Furthermore, the collaboration between MediaTek and Nvidia in the field of automotive artificial intelligence should broaden its artificial intelligence opportunities.

edit/ruby