The chip industry experienced a turbulent week, described as a week of 'two extremes,' starting with $ASML Holding (ASML.US)$ the performance disaster dragging down semiconductor stocks, followed by $Taiwan Semiconductor (TSM.US)$ The impressive financial report has reignited the market's optimistic sentiment towards the unchanging demand for AI.

Overnight, Taiwan Semiconductor soared nearly 10%, with a total market cap exceeding $1 trillion, making it the first Asian technology company to achieve a market cap of over $1 trillion in history.

What excites the market is that Taiwan Semiconductor shattered the market's pessimistic expectations with strong performance. During the conference call, Taiwan Semiconductor clearly stated that "AI demand is real". This statement directly ignited market enthusiasm, driving Nvidia to achieve a historic high.

Looking at the year-to-date increase, $NVIDIA (NVDA.US)$ Undoubtedly the dominant player in ai chip, with a staggering 177% increase in the year. $Arm Holdings (ARM.US)$ Closely following, with a cumulative increase of over 106%. $Taiwan Semiconductor (TSM.US)$ The stock price has also doubled, known as the "mini nvidia." $Astera Labs (ALAB.US)$ Also accumulated an increase of over 83%. $Broadcom (AVGO.US)$ The increase was as high as 64%.

How does Wall Street view the future of chip stocks?

The recent performances of ASML Holding and Taiwan Semiconductor, one cold and one hot, also reflect the internal differentiation of the chip industry.

Analysts believe that besides Taiwan Semiconductor, ASML Holding's customers, Intel and Samsung, are facing their own challenges. Meanwhile, the demand from Taiwan Semiconductor's customers like Nvidia remains strong, making Taiwan Semiconductor a better indicator of AI demand.

Therefore, the performance of these two core giants in the chip industry, ASML Holding and Taiwan Semiconductor, collectively shows that the stock logic support closely related to AI chips is extremely solid. The rising trend of leaders in AI chips like Nvidia may be far from over.

Morgan Stanley has also pointed out historically that the market cycles of logic chips and memory chips often synchronize. However, currently the two are gradually diverging in cycles. The demand for logic chips (such as AI chips) is being driven by the rapid development of AI technology, while the memory chip market (such as DRAM) is relatively weak due to supply-demand imbalances. Morgan Stanley emphasizes that AI technology is the main driving force behind the continuous growth of the logic chip market.

In contrast, Morgan Stanley believes that the memory chip market faces more challenges. The bank points out that the memory chip industry is being affected by oversupply and price declines, especially with more competitors entering the DRAM market, intensifying industry competition and entering a downward cycle.

In addition, $Advanced Micro Devices (AMD.US)$ CEO Su Zifeng recently mentioned at a new product launch event that the demand for datacenter AI chips, including AI GPUs, far exceeds expectations. It is estimated that by 2027, the market size of datacenter AI chips will reach $400 billion and further rise to $500 billion in 2028. This means that the compound annual growth rate of the global datacenter AI chip market is expected to exceed 60% from 2023 to 2028.

In a recent research report, Bank of America stated that the global artificial intelligence frenzy is still in its early stages, similar to the development path of the internet in the 1990s. It can be compared to the "1996 moment" of the booming internet, meaning that in the view of the Bank of America analysis team, the AI frenzy is still in a very early stage.

According to the latest forecast data from Citigroup, a financial giant on Wall Street, by 2025, the four largest tech giants in the USA - $Amazon (AMZN.US)$Please use your Futubull account to access the feature.$Alphabet-C (GOOG.US)$Please use your Futubull account to access the feature.$Microsoft (MSFT.US)$Please use your Futubull account to access the feature.$Meta Platforms (META.US)$ Datacenter capital expenditures are expected to increase by 40% to 50% year-on-year. The significant incremental spending by tech giants in datacenters is expected to drive their position as the "shovel sellers" in the AI infrastructure space. $NVIDIA (NVDA.US)$Please use your Futubull account to access the feature.$Arista Networks (ANET.US)$ Major datacenter networking technology giants' stock prices continue to be favored by international funds.

Especially in the datacenter AI chip field, the dominant player with a market share as high as 80%-90% - Nvidia may continuously set new historical highs in stock price, surpassing the analysts' widely expected $150 target price, which may just be a matter of time.

It is worth noting that Bank of America has raised its earnings per share forecast for Nvidia in the coming years, as the company holds 80%-85% market share in the potential $400 billion market. Bank of America reiterated a 'buy' rating for Nvidia, raising the target price from $165 to $190, representing a nearly 40% increase from the current level.

Editor/Somer