News
At TSMC’s earnings call on the 17th, the company revealed that its CoWoS (Chip-on-Wafer-on-Substrate) capacity will double each year in 2024 and 2025, but demand will continue to outpace supply. According to a report from Money DJ, the CoWoS expansion wave is expected to extend into 2026, promising strong growth for equipment suppliers for at least the next two to three years.
TSMC stated that advanced packaging currently accounts for approximately 7-9% of its revenue, and growth in this segment is expected to outpace the company’s average over the next five years. While the gross margin for advanced packaging is slightly below the company average, it is steadily approaching it. Regarding CoWoS capacity, customer demand significantly exceeds TSMC’s ability to supply, even with production capacity doubling year-on-year in both 2024 and 2025.
According to Money DJ, citing supply chain sources, TSMC has already provided equipment manufacturers with its machine requirements for 2026 and placed orders. Delivery schedules for next year are essentially fully booked, and TSMC is currently working with equipment suppliers to finalize shipment and installation plans for 2026.
The report noted that TSMC’s CoWoS monthly production capacity is expected to reach 35,000 to 40,000 wafers this year, and surge to 80,000 wafers per month next year. Originally, the expansion wave was anticipated to slow somewhat by 2026, with monthly capacity reaching around 100,000 to 120,000 wafers. However, strong and urgent demand from major AI customers continues to drive capacity needs, and with the addition of more equipment, TSMC’s CoWoS capacity could still see significant expansion, potentially reaching 140,000 to 150,000 wafers per month by 2026.
In addition, the report provided an overview of TSMC’s advanced packaging supply chain. Key suppliers for wet process equipment include GPTC and Scientech, which provide automated wet benches and single wafer spin processors. Scientech holds a significant share of CoWoS equipment orders, while GPTC remains a key global supplier for major packaging and testing companies like ASE, Micron, Amkor, and Chinese packaging firms.
(Photo credit: TSMC)
News
According to a report from Wccftech, Intel has stepped away from competing with NVIDIA in AI computing power and the market of training large-scale AI models. Instead, the company is now entering a less saturated segment of the AI market, focusing on its new cost-effective AI accelerator, Gaudi 3.
Intel aims to present Gaudi 3 as the product with the best price-to-performance ratio, though Gaudi 3 is “not catching up” to NVIDIA’s latest GPU from a head-to-head performance standpoint, citing the words from Anil Nanduri, head of Intel’s AI acceleration office, during an interview with CRN.
Nevertheless, Nanduri highlights that the Gaudi 3 accelerator chip is ideal for supporting cost-effective systems that run task-based and open-source models for enterprises.
On the other hand, according to the report from Wccftech, Intel claims that following the introduction of “reasoning-focused” LLM models, its Gaudi 3 lineup delivers performance comparable to NVIDIA’s well-known H100 AI accelerator, especially in inference workloads.
According to the report from CRN, Intel claims that Gaudi 3 is about 9% faster than H100 in the Llama 3 model and 80% more cost-effective; in the Llama 2 model, Gaudi 3 is 19% faster and the price-performance difference is up to 2 times.
However, when evaluated in terms of floating-point operations, the Gaudi 3 AI GPUs fall short compared to NVIDIA’s options, indicating that high-end AI performance isn’t currently Intel’s strength.
Therefore, according to the report from Wccftech, Intel has no plans to compete directly with NVIDIA’s GPUs. Furthermore, there is another rising competitor in the AI computing sector: AMD.
Regarding the reason behind Intel’s choice, the company believes that smaller LLM models will see increased acceptance as the initial excitement around AI and the buzz surrounding large-scale data centers diminish, as indicated by the report in Wccftech.
Read more
(Photo credit: Intel)
News
Rumors have been circulating for a while that Intel is considering to sell off its FPGA unit Altera, and later subsided as Altera reiterated that it still eyes an IPO in 2026. However, Intel’s stance of regarding Altera to be a critical part of its core business seems to waver, according to the latest report by CNBC.
Citing sources familiar with the situation, CNBC notes that Intel is seeking to sell at least a minority stake in Altera, a move that could generate several billion dollars in cash for the struggling semiconductor giant.
This week, the company was said to be reaching out to various private equity and strategic investors regarding Altera, according to the sources cited by the report. It is worth noting that Intel has indicated that acquiring a majority stake in the business is also a possibility.
A month ago, CEO Pat Gelsinger mentioned in a press release that Intel is working to carefully manage its cash as the company meaningfully improves its balance sheet and liquidity.
The efforts, in his own words, include selling part of Intel’s stake in Altera, something has been talked about publicly several times and has long been part of its strategy to generate proceeds for Intel on Altera’s path to an IPO.
According to CNBC, Intel aims to strike a deal that values Altera at around USD 17 billion, an amount approximately equivalent to the USD 16.7 billion Intel paid to acquire Altera in 2015.
An Intel representative declined to comment on the matter, according to CNBC.
Qualcomm, another tech giant who has expressed its interest in acquiring Intel, is said to be investigating the possibility of acquiring parts of Intel’s design business to enhance its product portfolio, but the decision might not be made after the U.S. presidential election, according to Bloomberg. Whether the development of the Altera sale would cause an impact to the potential deal remains to be seen.
Read more
(Photo credit: Intel)
News
During the earnings call on October 17th, TSMC revealed that electricity prices for its factories in Taiwan have doubled in recent years, making them the highest among all its global operations.
According to a report from CNA, TSMC stated that factors influencing next year’s gross margin include electricity costs, exchange rates, and the adoption of advanced nodes. The increase of overseas factories’s capacity might also reduce the gross margin by approximately 2 to 3 percentage points.
According to the report, TSMC’s Chief Financial Officer Wendell Huang has pointed out that the rise in electricity prices in Taiwan is one of the important factors affecting TSMC’s gross profit margin. He expects that rising electricity costs, coupled with other expense factors, will affect gross profit margin performance next year by at least 1%.
Huang said that now the electricity price of TSMC’s factories in Taiwan is the highest among all its operating factories in the world, and the electricity price adjustments in October this year has increased TSMC’s electricity charges by 14%.
According to the report, Huang mentioned that in the past few years, the electricty price for TSMC in Taiwan has doubled. In 2022, the electricity price increased by 15% and in 2023 by 17%. In the first half of this year, there was an increase of 25%.
Recently, TSMC’s increasing power consumption has sparked concerns regarding Taiwan’s energy supply. Citing a report by S&P, a report by Wccftech highlights that compared with 2023, the foundry giant’s electricity consumption could nearly triple by 2030, accounting for about 24% of the island’s total electricity usage.
According to Wccftech, TSMC’s shift to 3nm chip production is driving S&P’s forecasts for the company’s soaring electricity consumption. On the other hand, the report also cites data from Taiwan’s state-owned electricity provider, TaiPower, to show that the island’s electricity reserve margin continues to fall short of the government’s 15% target.
Other factors that might influence the gross margin next year are also discussed in the earnings call. According to the report, the shift towards more advanced nodes, such as 3nm and 2nm, has significantly impacted the gross margin.
TSMC has transitioned some of its capacity from 5nm to 3nm chip production to meet the strong demand for 3nm chips. The company also plans to begin 2nm production in 2026. All of these factors contribute to increased costs and affect the gross margin.
Huang also noted that fluctuations in the exchange rate could affect the gross margin next year. According to the report, industry insiders predict a roughly 1% change in the USD to NTD exchange rate, which could impact TSMC’s gross margin by approximately 0.4 percentage points.
In addition, during the earnings call, institutional investors expressed concerns about antitrust issues, according to the report. Chairman and CEO C.C. Wei highlighted that TSMC not only manufactures wafers but also engages in advanced packaging, testing, and mask production. Revenue from these additional sectors contributes approximately 10% to the overall revenue.
According to the report, Wei stated that when including advanced packaging, testing, mask production, and other projects, TSMC holds a market share of approximately 30% in the semiconductor wafer industry. He emphasized that the company does not have a monopoly and therefore does not face antitrust issues.
Read more
(Photo credit: TSMC)
Insights
China’s monthly economic data in September showed signs of improvement, according to China’s National Bureau of Statistics on October 18.
In terms of consumption, retail sales grew by 3.2% year-over-year in September, up by 1.1 percentage points from the previous month and exceeding market expectations of 2.5%. This growth was mainly driven by household electronics, which surged by 20.5% year-over-year, up 17.1 percentage points from the previous month, reflecting the continued impact of China’s “trade-in” policy. Auto sales in September rose by 0.4% year-over-year, an increase of 7.7 percentage points from the previous month, reflecting the industry’s entry into the peak sales season of “golden September, silver October.”
In the industrial sector, industrial output increased by 5.4% year-over-year in September, an improvement of 1.3 percentage points from the previous month, and surpassing market expectations of 4.6%. High-tech manufacturing continued to drive overall industrial growth, with a year-over-year increase of 10.1%, up 1.5 percentage points from the previous month, reflecting China’s focus on high-quality development and new productivity policies.
In terms of investment, cumulative fixed-asset investment grew by 3.4% year-over-year in September, on par with the previous month, and slightly above market expectations of 3.3%. Industrial investment increased by 12.3%, up 0.1 percentage points from the previous period, while infrastructure investment grew by 4.1%, down 0.3 percentage points from the previous period. Additionally, both private and foreign investment continued to decline, signaling weaker business confidence in future prospects.
Overall, the September data suggests early signs of improvement in China’s domestic economy. However, the latest GDP data reveals that real GDP grew by 4.6% year-over-year in the third quarter, down by 0.1 percentage points from the previous quarter. Cumulative GDP growth for the first three quarters stood at 4.8%, still below the annual target of 5%.
Despite the Chinese government’s announcement of a series of monetary easing measures in late September, the ongoing slowdown in economic growth indicates that the recovery is still in its early stages. The government will need to expedite the implementation of large-scale fiscal policies to further stimulate economic growth.