The world’s four major CSPs (Cloud Service Providers) – Microsoft, Google, Amazon, and META – are continuously expanding their AI infrastructure, with their combined capital expenditures projected to reach USD 170 billion this year. According to the industry sources cited in a report from Commercial Times, it’s pointed out that due to the surge in demand for AI chips and the increased area of silicon interposers, the number of chips that can be produced from a single 12-inch wafer is decreasing. This situation is expected to cause the CoWoS (Chip on Wafer on Substrate) production capacity under TSMC to remain in short supply.

Regarding CoWoS, according to TrendForce, the introduction of NVIDIA’s B series, including GB200, B100, B200, is expected to consume more CoWoS production capacity. TSMC has also increased its demand for CoWoS production capacity for the entire year of 2024, with estimated monthly capacity approaching 40,000 by the year-end, compared to an increase of over 150% from the total capacity in 2023. A possibility exists for the total production capacity to nearly double in 2025.

However, with NVIDIA releasing the B100 and B200, the interposer area used by a single chip will be larger than before, meaning the number of interposers obtained from a 12-inch wafer will further decrease, resulting in CoWoS production capacity being unable to meet GPU demand. Meanwhile, the number of HBM units installed is also multiplying.

Moreover, in CoWoS, multiple HBMs are placed around the GPU, and HBMs are also considered one of the bottlenecks. Industry sources indicate that HBM is a significant challenge, with the number of EUV (Extreme Ultraviolet Lithography) layers gradually increasing. For example, SK Hynix, which holds the leading market share in HBM, applied a single EUV layer during its 1α production phase. Starting this year, the company is transitioning to 1β, potentially increasing the application of EUV by three to four times.

In addition to the increased technical difficulty, the number of DRAM units within HBM  has also increased with each iteration. The number of DRAMs stacked in HBM2 ranges from 4 to 8, while HBM3/3e increases this to 8 to 12, and HBM4 will further raise the number of stacked DRAMs to 16.

Given these dual bottlenecks, overcoming these challenges in the short term remains difficult. Competitors are also proposing solutions; for instance, Intel is using rectangular glass substrates to replace 12-inch wafer interposers. However, this approach requires significant preparation, time, and research and development investment, and breakthroughs from industry players are still awaited.

Read more

• [News] TSMC’s Advanced Packaging Capacity Fully Booked by NVIDIA and AMD Through Next Year
• [Insights] Big Four CSPs Continue to Shine in Q1 2024 Financial Reports, AI Returns Garnering Attention

(Photo credit: NVIDIA)

As Apple keeps advancing in AI as well as developing its own in-house processors, industry sources indicated that the tech giant’s Chief Operating Officer (COO) Jeff Williams recently made a visit to TSMC, and was personally received by TSMC’s President, C.C. Wei, according a report by Economic Daily News.

The low-profile visit was made to secure TSMC’s advanced manufacturing capacity, potentially 2nm process, booked for Apple’s in-house AI-chips, according to the report.

Apple has been collaborating with TSMC for many years on the A-series processors used in iPhones. In recent years, Apple initiated the long-term Apple Silicon project, creating the M-series processors for MacBook and iPad, with Williams playing a key role. Thus, his recent visit to Taiwan has garnered significant industry attention.

Apple did not respond to the rumor. TSMC, on the other hand, has maintained its usual stance, not commenting on market speculations related to specific customers.

According to an earlier report from The Wallstreet Journal, Apple has been working closely with TSMC to design and produce its own AI chips tailored for data centers in the primary stage. It is suggested that Apple’s server chips may focus on executing AI models, particularly in AI inference, rather than AI training, where NVIDIA’s chips currently dominate.

Also, in a bid to seize the AI PC market opportunity, Apple’s new iPad Pro launched in early May has featured its in-house M4 chip. In an earlier report by Wccftech, Apple’s M4 chip adopts TSMC’s N3E process, aligning with Apple’s plans for a major performance upgrade for Mac.

In addition to Apple, with the flourishing of AI applications, TSMC has also reportedly beening working closely with the other two major AI giants, NVIDIA and AMD. It’s reported by the Economic Daily News that they have secured TSMC’s advanced packaging capacity for CoWoS and SoIC packaging through this year and the next, bolstering TSMC’s AI-related business orders.

Read more

• [News] Apple M4 Incoming, Boosting TSMC’s 3nm Production
• [News] Apple Unveiled M4 Chip for AI, Heralding a New Era of AI PC

(Photo credit: TSMC)

In addition to the aggressive overseas expansion plans recently, TSMC also demonstrates its ambition of increasing specialty capacity, targeting to be expanded by 50% by 2027, according to a report by AnandTech. A key driver of this demand will be TSMC’s forthcoming specialty node, N4e, a 4nm-class ultra-low-power production node.

Citing Kevin Zhang, TSMC’s Senior Vice President of the Business Development and Overseas Operations Office, the report revealed that TSMC plans to expand its specialty capacity by up to 1.5 times in the next four to five years. To accomplish this goal, it would not only convert existing capacity, but construct new fab space dedicated to specialty processes.

TSMC offers a range of specialty nodes catering to various applications such as power semiconductors, mixed analog I/O, and ultra-low-power applications (e.g., IoT), according to the report. Currently, the semiconductor giant’s most advanced specialty node is N6e, a variant derived from N7/N6 that accommodates operating voltages ranging from 0.4V to 0.9V. With N4e, TSMC aims to support voltages below 0.4V.

According to the materials TSMC provided in its latest earnings call, in the first quarter, HPC accounted for 46% of its total revenue, while IoT-related and automotive applications accounted for 6% of its total revenue, respectively. All the applications mentioned above are closely connected to specialty nodes.

TSMC’s overseas expansion plans are also closely related to its focus on specialty nodes. At the grand opening of JASM’s first Kumamoto plant in February, TSMC Chairman Mark Liu stated that JASM would use the latest green manufacturing practices to produce best-in-class specialty semiconductor technology.

Read more

• [News] TSMC Reportedly Prepares Next-generation HBM4 Manufacturing, Utilizing 12nm and 5nm Process Nodes

(Photo credit: TSMC)

With the emergence of net-zero carbon emissions and generative AI, the rising electricity demand will also become a major, long-term global trend accompanying these two hotly debated issues.

For many governments and corporations, satisfying their electricity demand while meeting policy mandates related to low carbon emissions or net-zero carbon emissions is now a huge challenge.

According to data from the International Energy Agency (IEA), the total electricity consumption of data centers worldwide is forecasted to surpass 1,000 terawatt-hours by 2026. This amount is equivalent to the annual electricity consumption of Japan. Additionally, with the electrification of vehicles and decarbonization of industries being irreversible long-term trends, electric vehicle sales are forecasted to account for 67% of total car sales by 2030. Hence, these projections indicate that global electricity consumption will continue to increase in the future.

TrendForce analyst Danis He stated that in the future, the share of renewable energy in electricity generation will continue to expand. However, no single generation technology can meet much of the electricity demand. Therefore, the ultimate goal of green energy transition is to establish a “stable, low-carbon power infrastructure” that requires multiple sources of energy technologies working in sync.

So, in terms of power infrastructure planning, how can governments and businesses simultaneously meet the rising electricity demand while also meeting requirements related to low carbon emissions and non-carbon energy?

He pointed out that currently, the costs of solar photovoltaics and wind power have significantly dropped, making them even more competitive against traditional coal-fired power stations. Consequently, many countries have been able to scale up their deployment of renewable generation systems and advance numerous clean energy projects.

Even though solar photovoltaics and wind power are intermittent energy sources, they can play a vital role in stabilizing the operation of a power grid by coupling with lithium-ion battery energy storage systems, the costs of which have also fallen considerably in recent years. Additionally, redox flow batteries are also an option, especially for applications that require long-term energy storage.

Global Top 10 Photovoltaics Companies Are Based either in China or the United States as They Control Key Technologies and Huge Market Shares

Photovoltaic systems generally have relatively low installation requirements and will produce electricity as long as they are exposed to sunlight. Hence, they are widely deployed for various applications such as rooftop solar, ground-mounted photovoltaic installations, and floating photovoltaic installations. He stated that some manufacturers of photovoltaic products have already established carbon-neutral manufacturing processes. Under this premise, solar photovoltaics can be regarded as being far less polluting compared with other energy sources.

Presently, there are two types of silicon wafers used in the manufacturing of photovoltaic cells: N-type and P-type. N-type tunnel oxide passivated contact (TOPCon) cells have emerged as the market mainstream, and their demand is expected to soar in 2024. Approximately 57% of the total worldwide production capacity for photovoltaic cells is allocated to N-type TOPCon cells. P-type passivated emitter rear contact (PERC) cells have secured second place with around 32% of the total production capacity.

He asserted that in terms of rankings of cell manufacturers based on production capacity and shipment volume, the trend of the strong getting stronger is increasingly evident. The market shares of the leading manufacturers have remained roughly the same compared to 2023. While the rankings of the top 10 have seen some minor changes, the order has generally been constant. This group includes companies such as Canadian Solar, LONGi, Jinko Solar, Trina Solar, and JA Solar, all based in China, as well as US companies such as First Solar.

In recent years, the government of Taiwan has been actively promoting initiatives related to green energy transition. Major photovoltaic companies in Taiwan include TSEC, URE, Motech, and Anji Technology. Currently, they are expanding their presence from the upstream sections of the supply chain, such as photovoltaic cells and modules, to the downstream sections, such as investments in photovoltaic power stations.

Synergy of Various Low-carbon Energy Sources Has Revealed a Promising Future for Hydrogen and CCUS

There are countless types of renewable energy available today, including many that are specific to different environments and geographical conditions, such as tidal power and geothermal energy. These energy sources primarily target niche markets, and widespread commercialization is still far off for them. However, for countries that possess abundant geothermal resources or other geographical advantages, such as being surrounded by the ocean, they ultimately have ample natural resources that can be utilized for renewable generation.

Since geographical environments differ from region to region, the deployment of electricity generation systems must consider various factors in order to realize the full-scale adoption of low-carbon energy. Nevertheless, several newly emerged energy sources have generation costs that are still significantly higher compared to traditional fossil fuels.

Among these emerging energy sources, green hydrogen is highly anticipated to play a critical role in achieving net-zero carbon emissions. He said that while green hydrogen is very costly at this moment, it will undoubtedly become mainstream in the next decade or two, serving as a promising low-carbon energy source and an option for energy storage beyond 2030.

According to the IEA’s Global Hydrogen Review 2023, global hydrogen demand is forecasted to reach 150 million tons by 2030, and the total installed capacity of generation systems that consume either hydrogen or ammonia will exceed 5.8 gigawatts. Furthermore, this report states that in order to meet the 2050 net-zero target, hydrogen energy needs to account for at least 13% of the overall energy consumption.

The use of hydrogen fuel for transportation has been advocated by companies such as Air Liquide from France. Apart from collaborating with Toyota from Japan and Hyundai from South Korea, Air Liquide is also promoting the adoption of green hydrogen in Taiwan. Another example is the US-based GE partnering with TWAIDC to develop a generation unit that runs on both natural gas and hydrogen.

As for carbon capture, utilization, and storage (CCUS), cost is also a challenge at the current stage. In order to ensure a stable energy supply, the presence of thermal power stations that burn fossil fuels remains necessary, though these facilities need to undergo further transformation. He pointed out that against the backdrop of surging electricity demand, countries around the world are vigorously advancing towards renewable energy as they plan and develop power infrastructure. However, at the same time, new thermal generation units featuring some type of CCUS technology are also being developed.

CCUS has undergone several decades of development and is expected to become an indispensable part of our future energy system. In 2021, a Swiss startup named Climeworks activated the world’s first facility for carbon capture and conversion. Capable of capturing 4,000 tons of carbon dioxide annually, this facility represents a milestone in the development of an industry for direct carbon capture.

Also, in 2023, Microsoft signed a contract to buy “carbon removal credits” from another US company aptly named CarbonCapture. Having developed a process that removes carbon dioxide from ambient air and stores it underground, CarbonCapture assists Microsoft in eliminating its historical carbon emissions and achieving its carbon-negative goals.

Looking at Taiwan, major local producers of carbon emissions such as Formosa Plastics, Taipower, and CSC are now actively investing in the development of technologies related to CCUS. However, He said that at the current stage, implementing CCUS to reduce carbon footprints can incur a substantial cost and affect the efficiency of power generation. Studies have shown that coal-fired power stations that have adopted CCUS have seen a 60% increase in generation cost as well as a 20-30% reduction in generation efficiency. Moreover, Taiwan cannot replicate the arbitrary carbon storage methods used in the US and Europe due to its limited landmass. Without sufficient subsidies, CCUS cannot be applied on a large scale.

Integration of AI into Smart Grids Will Massively Boost Efficiency in Electricity Usage

As countries gradually proceed with the green energy transition, the market penetration rate of intermittent renewable energy sources climbs, while traditional thermal baseload power stations shut down. This trend has the effect of heightening uncertainty in the operation of power grids. Therefore, many new kinds of methods and energy sources have been developed to regulate the grid. Furthermore, the entire electric power system is inevitably advancing into the era of smart grids.

The purpose of the smart grid is to reduce electricity loss and effectively integrate renewable energy. To this end, it is anticipated that technologies related to artificial intelligence (AI) will have an increasing presence in the field of grid operation. Presently, several Taiwan-based companies are involved in the downstream of the industry chain for smart grid, providing services for various applications. Among these players, Delta Electronics, Chung-Hsin Electric & Machinery, HDRE, Advantech, and Tatung System Technologies all provide system integration services.

Furthermore, their solutions can leverage big data to improve generation efficiency and perform predictive analysis. For instance, by collecting and processing data related to weather forecasts and performances of individual generation units, an AI-enabled solution can predict the amount of electricity produced by a solar or wind project. Furthermore, these companies also provide hardware and software that enable real-time monitoring of the grid, providing indicators such as the quality of electricity supply, load, etc. Overall, their services and products can facilitate real-time power dispatching and optimization of the whole grid system.

Currently, AI cannot be extensively applied for grid scheduling. As He pointed out, in terms of legal responsibilities and obligations, if errors occur in power dispatching, it will be difficult to determine accountability. No sector within the entire electric power system are able to provide sufficient safety assurances or guarantees when it comes to adopting AI for certain applications. All in all, this is one of the major challenges in the future development of the smart grid.

Join the AI grand event at Computex 2024, alongside CEOs from AMD, Intel, Qualcomm, and ARM. Discover more about this expo! https://bit.ly/44Gm0pK

(Photo credit: Delta)

Previously, the U.S. Department of Commerce revoked Intel and Qualcomm’s export licenses to Huawei, leading to speculation that they are now prohibited from collaborating with Huawei. According to a report from TechNews, Qualcomm anticipates that after 2024, it will no longer receive product revenue from Huawei but will continue to collect patent royalties.

As per a report from global media outlet tom’s Hardware, on May 7, 2024, the U.S. Department of Commerce informed Qualcomm that it was revoking the company’s license to export 4G and certain other integrated circuit products, including Wi-Fi products, to Huawei. a nd its affiliates and subsidiaries, effective immediately.Consequently, Qualcomm expects no product revenue from Huawei after this year.

According to TechNews, while Qualcomm used to provide processors to Huawei for use in its smartphones, Huawei’s HiSilicon division has developed its own chipsets, Kirin 9000 and 9010, therefore barely needing the support from Snapdragon processors.

Qualcomm reportedly noted in another statement that Huawei has recently launched new 5G-supported devices using its own IC products. Although Qualcomm can still sell IC products to Huawei under the current license, it does not expect to receive any product revenue from Huawei after this year.

Despite having its own processors, Huawei lacks the alternative for Intel’s Core or Xeon CPUs from PCs and servers, and will likely continue using them for the foreseeable future, according to tom’s Hardware. Meanwhile, Qualcomm may continue to collect patent royalties from Huawei and other Chinese smartphone manufacturers.

Qualcomm also mentioned that it has recently extended, renewed, or signed licensing agreements with several major OEMs. Negotiations are ongoing with key OEMs, including Huawei, for agreements set to expire at the beginning of fiscal year 2025.

Read more

• [News] As U.S. Export Restrictions Tighten, Qualcomm Reportedly Affirms Huawei May no Longer Need its Chips
• [News] Qualcomm Advances into AI PC Market, Reportedly Adopting 4nm Process

(Photo credit: Qualcomm)