Fortunes Diverge for Semiconductor Companies! Chip Stocks Soar Thanks to AI

Analysts Notebook joined discussion · Oct 22 07:19

The global semiconductor industry, valued at $530 billion, is undergoing a transformative shift driven by the rapid rise of AI. Companies that have embraced AI are thriving, while those that have missed this wave are struggling. Based on recent analyses, this divide is only expected to grow wider.

Fortunes Diverge for Semiconductor Companies! Chip Stocks Soar Thanks to AI

The AI Wave: Driving the Industry Split

AI is fundamentally reshaping the semiconductor landscape. Tech giants like $Microsoft (MSFT.US)$, $Alphabet-C (GOOG.US)$, $Amazon (AMZN.US)$, and $Meta Platforms (META.US)$ are investing massive capital in AI infrastructure. According to Solita Marcelli, Chief Investment Officer at UBS Global Wealth Management, AI-related semiconductor sales will reach $168 billion in 2023 and surge to $245 billion by 2025.

One of the biggest beneficiaries of this trend is $NVIDIA (NVDA.US)$, whose AI accelerator chips dominate the market. These chips are widely used in data centers and high-performance computing. Nvidia's stock has surged by over 175% in 2024, pushing its market value to nearly $3.4 trillion. Nvidia's CEO, Jensen Huang, recently confirmed that the company's latest Blackwell chip is in full production, with strong customer demand.

$Taiwan Semiconductor (TSM.US)$ is another major beneficiary, as it produces chips for companies like Nvidia and Apple. TSMC's CEO, C.C. Wei, recently stated that AI-driven demand is real and that overall chip demand has stabilized, showing signs of improvement.

$Advanced Micro Devices (AMD.US)$ and $Broadcom (AVGO.US)$ are two other companies gaining ground in the AI space. AMD is developing new AI accelerator chips to challenge Nvidia's dominance, while Broadcom's custom chips and networking semiconductors are crucial for data centers. Despite a recent drop in Broadcom's stock due to weaker-than-expected non-AI business results, analysts believe the company is well-positioned for future growth in AI-related markets.

Many analysts believe that AI will continue to impact the semiconductor industry for a long time.Ryuta Makino, a research analyst at Gabelli Funds, expects AI-driven demand to lead the industry's growth for several years. In contrast, non-AI chip demand may only recover once the global economy stabilizes. This divergence will likely persist through 2025, with AI remaining the primary growth driver.

The Struggles of Traditional Semiconductor Giants

While AI-related businesses thrive, many traditional chipmakers face significant challenges due to weak demand in non-AI sectors. $ASML Holding (ASML.US)$ recently slashed its 2025 sales forecast due to a sharp decline in orders outside AI. ASML's CEO, Christophe Fouquet, warned that the semiconductor market would be "bleak without AI."

Other industry giants like Intel and Samsung are also struggling. Intel has been cutting costs and delaying new factory construction to cope with declining sales and mounting losses. Meanwhile, Samsung faced significant setbacks due to delays in producing its high-bandwidth memory chips, leading to disappointing Q3 results.

Challenges for Semiconductor Equipment Makers

Chip equipment suppliers are also feeling the effects of this industry shift. Companies like ASML, $Applied Materials (AMAT.US)$, and $Lam Research (LRCX.US)$ have recently seen their stock prices plummet. Investors are concerned about slowing demand for chip-making equipment, particularly in non-AI sectors. CJ Muse, an analyst at Cantor Fitzgerald, noted that while long-term leaders like ASML were expected to perform well, their results have fallen short, signaling broader industry challenges.

Before ASML's disappointing earnings report, many investors believed a new "boom cycle" for the semiconductor industry had already begun in 2024.

However, despite the high demand for NVIDIA's AI GPUs from 2023 to 2024, ASML's reports showed a different story. TSMC kept buying ASML's advanced EUV machines, but not in the large amounts analysts had predicted. By the third quarter, TSMC even slowed its purchases. This slowdown shows that demand for non-AI chips is still weak, and the AI boom alone has yet to push the semiconductor industry into growth. As a result, some chip factories now have more production capacity than needed.

Since semiconductor chips are heavily used in consumer electronics like PCs and smartphones, the industry tends to follow a cyclical pattern closely tied to overall economic conditions. The chart clearly shows these cycles, with growth phases usually lasting 1.5 to 2 years and full inventory cycles taking 3 to 4 years to complete.

Bernstein analysts warned in a recent report: "The recovery in chip demand may take longer than expected, which will delay factory expansions. This is a key issue we'll face in 2025." They added, "We may need to be patient until the industry's recovery becomes clearer."

Jefferies analyst Janardan Menon also stated in a report: "ASML's earnings reflect that although demand for AI-related chips remains very strong, recovery in other areas has been unusually slow, and this trend may continue into 2025."

Disclaimer: Moomoo Technologies Inc. is providing this content for information and educational use only. Read more