According to Zhitong Finance APP, due to the lack of technological leadership in the artificial intelligence chip field and the tiny market share in this area, Intel (INTC.US) and Samsung Electronics have evaporated a total of approximately $227 billion in market value this year in the unprecedented AI frenzy, which is close to the total market value of another chip giant, AMD (AMD.US).

Although the unexpected rebound in earnings reports of the two major chip giants this week has provided immediate relief for their persistently low stock prices, investment institutions and analysts generally believe that the fundamentals of these two chip giants have not undergone a significant change, they still have a long way to go, and must find ways to continuously gain market share in the artificial intelligence infrastructure field in order to completely reverse their weak fundamentals. The same goes for global chip companies, achieving high growth in performance or substantial reversal in fundamentals, must firmly grasp this AI frenzy and find ways to leverage the AI heat.

The American traditional chip company behind the once omnipresent 'Intel Inside' label globally, when entering the foundry business, paid an extremely high price, with huge losses in the foundry business for multiple quarters, and Nvidia, AMD, and Broadcom etc. chip giants continue to ignore Intel's strong promotion of foundry business, firmly sending huge AI chip foundry orders to TSMC, factors that have led the market to doubt Intel's 'chip foundry dream' may not be able to step out of the dream into reality.

More importantly, with Intel's longtime competitor AMD going all out to layout data center AI chips, an unimaginably vast and hottest global technology sector, Intel's long-cherished enterprise-level data center CPUs are starting to be neglected, forcing Intel to compete with AI chip dominator Nvidia. However, the fact proves that Intel's AI chips are struggling to shake Nvidia's and AMD's market share.

South Korean television, smart phone, and storage chip giant Samsung is struggling for its own dream of advanced process chip foundry of 3nm and below, falling significantly behind in artificial intelligence storage systems, especially in HBM storage systems tightly integrated with Nvidia's high-performance AI GPUs, compared to Samsung's longtime competitor SK Hynix also headquartered in South Korea.

It is understood that Samsung has acknowledged in its latest earnings report that its top-profit, most advanced HBM3E series storage systems have experienced unexpected delays in obtaining certifications from major clients like Nvidia, while SK Hynix confirmed last week that they plan to supply their 12-layer HBM3E to clients including Nvidia starting in the fourth quarter. Apart from lagging behind in HBM, Samsung has made almost no progress in the outsourced production of custom AI chips compared to the foundry king TSMC.

taiwan semiconductor and SK hynix will comprehensively defeat intel and samsung in 2024.

Chip companies may need to find ways to ride the wave of AI popularity in order to achieve strong performance growth.

According to Wong Kok Hoong, the institution stock sales and trading director from Maybank Securities Pte., the gap between winners and losers in the field of artificial intelligence "will not narrow quickly," and although intel's third-quarter sales data exceeded expectations, its sales of artificial intelligence-related chips fell far short of expectations. Overall, artificial intelligence will be an extremely important field in the next five to ten years.

AMD CEO Lisa Su recently stated at a new product launch event that the demand for datacenter ai chips, including AI GPU, far exceeds expectations. It is expected that by 2027, the datacenter AI chip market size will reach $400 billion, further rising to $500 billion in 2028. This implies that the compound annual growth rate of the global datacenter AI chip market size is expected to exceed 60% from 2023 to 2028.

Artificial intelligence has completely overturned the valuation level of global chip companies. Samsung's total market cap is now not even a third of Taiwan Semiconductor, which holds the title of "King of Chip Manufacturing," the core chip outsourcing manufacturer for nvidia and AMD, the two top global AI chip giants. Just four years ago, Samsung and taiwan semiconductor were on similar levels in terms of market cap. intel, once the world's largest chip giant, is a globally highest market cap chip company that integrates chip design, manufacturing, packaging, and testing into one. However, taiwan semiconductor's current market cap is almost 10 times that of intel.

Following the announcement of the financial report after the post-market trading on Thursday, intel's stock price soared in post-market trading in the US, rising over 6% as of pre-market trading on Friday. The strong performance and sales prospects of this veteran chip giant have fueled investors' optimism, believing that the giant has the ability to regain some lost PC market share and slightly erode the market share of the AI chip dominator nvidia by up to 90%. Intel CEO Pat Gelsinger stated that the company has recently made a lot of positive changes through restructuring, although not all investors believe.

Intel's latest financial report data shows that third-quarter revenue fell by 6% to $13.3 billion, surpassing analysts' widespread expectation of $13 billion. Importantly, looking ahead to the fourth quarter, the California-based chip giant expects overall revenue to reach $13.3 billion to $14.3 billion, with analysts' average expectation of $13.6 billion; the giant also expects earnings per share of $0.12, while analysts' average expected earnings per share is $0.06.

"Intel is still a shadow of its former self," wrote Adam Crisafulli, founder and analyst of Vital Knowledge, in a report. "During the earnings conference call, the company admitted that its Gaudi ai chip would not achieve this year's $0.5 billion sales target, a data point that broadly underscores how the company is essentially a 'non-entity' participant in the current AI frenzy."

On Thursday, Samsung also provided some cheers as it cited 'meaningful' progress in supply qualification testing of its most advanced HBM storage system with major customers (presumed to be Nvidia). After the news broke, Samsung's stock price soared in the Seoul stock market but eventually settled unchanged by the close, essentially erasing gains. On the downside, the company's semiconductor business failed to meet market profit expectations, facing issues of traditional memory chip oversupply in the Chinese market, intense competition with Micron and SK Hynix in the high-end memory chip segment for consumer electronics, as well as escalating competition in enterprise-grade storage, like SSDs for data center companies.

Portfolio manager Gary Tan from Allspring Global Investments stated that the company's comments on the supply qualification testing of HBM storage system should provide some short-term relief to its stock price, but a long-term improvement will require the company to make difficult capital allocation decisions among its various business units and fully pivot towards the emerging field of HBM closely related to AI.

After Samsung released its latest financial data, analysts from Wall Street financial giants like Goldman Sachs and JPMorgan significantly reduced their target stock prices for this South Korean memory chip giant.

Broadcom, ASML Holding, and Taiwan Semiconductor have proven with actual performance: only AI can drive companies into a performance expansion curve.

Portfolio manager Jian Shi Cortesi from Gam Investment Management in Zurich stated that on the other side of the vast AI divide, recent incredibly strong performance results from Taiwan Semiconductor and SK Hynix confirm that they are the core beneficiaries in the Asian AI field. He added that the investment fund continues to hold these two chip companies. "AI is a theme that will continue for many years, and we are still in the very early stages," emphasized the investment manager.

Taiwan Semiconductor has significantly raised its overall revenue expectations for this year and the management of Taiwan Semiconductor has shown extremely optimistic sentiment towards revenue related to AI chips. Meanwhile, SK Hynix reported record operating profit and plans to launch the latest HBM storage system this quarter, with its technical progress in the HBM field far ahead of Samsung.

According to data compiled by institutions, this year Wall Street analysts have nearly doubled the target price for Taiwan Semiconductor within the next 12 months and raised the target price for SK Hynix by up to 60%. In contrast, analysts have significantly lowered the target price for Intel by 40% and around 5% for Samsung.

In fact, before Intel and Samsung released their latest financial reports, Wall Street analysts had reached a logical consensus on investing in chip stocks: only by closely embracing artificial intelligence can chip companies drive into a curve of earnings growth, and the higher the correlation between business and AI, the more likely the fundamentals are to show a booming growth trend.

One of the biggest winners in the global AI boom, Broadcom (AVGO.US), in the financial results for the third quarter ended on August 4th announced in early September, showed that the demand for chips related to AI is surging, while non-AI areas are still in a "darkest hour". Without AI contributions, the company's performance would be very bleak.

The customized AI chips and Ethernet chips provided by Broadcom are mainly used in major data centers worldwide, but in September, its stock price plummeted due to disappointing performance in a part of its business unrelated to artificial intelligence, and the latest performance forecast from Broadcom basically indicates that the growth rate of Broadcom's non-AI business is much slower than the market's expected recovery pace.

The third-quarter performance of the two core manufacturing forces in the global chip industry – ASML and Taiwan Semiconductor – showed that in this highest-end global industry chain worth as much as $530 billion, the actual performance gap between chip companies riding the unprecedented AI frenzy and those that have not fully caught up with this AI boom is getting bigger. It highlights that the market is betting heavily on listed chip companies related to AI, rather than those companies that mainly link their core business to non-AI areas.

"If there is no artificial intelligence, the entire chip market will be very difficult." Christophe Fouquet, CEO of Dutch lithography machine manufacturer ASML, said during the third-quarter earnings conference call. ASML's Q3 performance unexpectedly fell short, mainly due to continued softness in all chip demand outside of artificial intelligence, leading to a downward revision of the full-year sales forecast for 2025. ASML's Q3 orders fell significantly short of market expectations due to languishing demand in non-AI areas.

ASML's unexpected performance shortfall report has unveiled the latest dynamics in the global chip industry, namely: the AI boom is still unfolding, especially the demand for all types of AI chips focusing on B-end data centers remains very hot. However, in areas not related to AI, such as electric vehicles, industrial sectors, IoT devices, and a wide range of consumer electronics products, chip demand still remains weak or even significantly declining.

After chip manufacturing giant Taiwan Semiconductor significantly raised revenue expectations for 2024 and management emphasized the extremely crazy demand for AI, it greatly eased market concerns about the demand for the entire chip industry, and greatly boosted market optimism about the demand for AI chips. Apple, as well as AI chip giant NVIDIA, and Broadcom are all core chip foundry customers of Taiwan Semiconductor. Taiwan Semiconductor's CEO C.C. Wei stated in the earnings conference that industry growth is being strongly driven by AI-related factors and overall chip demand has become "stable", showing signs of improvement starting. When talking about market demand for AI chips, Taiwan Semiconductor's leader C.C. Wei emphasized in the earnings conference that the prospects for demand for AI chips are very optimistic, and stressed that the demand from Taiwan Semiconductor's customers for advanced CoWoS packaging far exceeds the company's supply.

"The company will fully address customers' demand for CoWoS advanced packaging capacity. Even if the capacity doubles this year and continues to double next year, it will still be far from enough," Wei Zhejia stated during the earnings conference. The capacity for CoWoS advanced packaging is crucial for a broader range of AI chips like Nvidia's Blackwell AI GPU. Nearly all AI innovators collaborate with Taiwan Semiconductor, confirming the real existence of AI-related demand, and I believe this is just the beginning.

Asml Holding and Taiwan Semiconductor, the two core giants of the semiconductor industry chain, have jointly demonstrated that the global AI spending wave is still unstoppable. Stocks closely related to AI chips have extremely strong support logic, and the stock price of leaders in AI chips such as Nvidia may continue to rise. Especially in the data center AI chip field, Nvidia, the AI chip leader with a market share as high as 80%-90%, may continue to set historical highs in stock prices, breaking through the widely expected $150 by Wall Street analysts, which may only be a matter of time.

Recently, the stock prices of semiconductor equipment companies have significantly underperformed the Philadelphia Semiconductor Index. Rarely, giants like Asml Holding, Kla Corp, and Lam Research in the semiconductor equipment field have been included in the "lagging list," while chip companies closely related to AI like Taiwan Semiconductor, Nvidia, and Broadcom have seen significant increases in stock prices. Semiconductor equipment companies provide chip manufacturing equipment focusing on core chip manufacturing processes such as lithography, etching, thin film deposition, multilayer interconnection, and thermal management. Therefore, compared to chip companies heavily benefiting from the global AI frenzy like Taiwan Semiconductor, Nvidia, AMD, and Broadcom, which are mostly related to AI chip demands, semiconductor equipment companies focusing on heavy assets still heavily rely on the overall chip industry demand rather than just the demand associated with AI.

"We expect this divergence to persist, as assuming all this demand is AI-driven is completely accurate." Senior analyst Ryuta Makino from Gabelli Funds expressed that he expects this significant divergence path to continue at least until 2025. Senior portfolio strategist Tim Ghriskey from Ingalls & Snyder recently stated that AI will be a long-term market focus.