Supermicro, a major server manufacturer, is facing potential delisting after failing to submit its financial report on time. Previously, the Economic Daily News reported that the company had halted expansion of its Malaysia plant. However, in response to inquiries from the Central News Agency, Supermicro clarified that its plans for the Malaysia facility remain unchanged, with production set to commence by the end of this year, and emphasized that customer orders are “still robust.”

According to the Economic Daily News, Supermicro’s Malaysia plant was originally scheduled to come online in the fourth quarter, doubling the company’s production capacity. However, amid financial turmoil, the project has been temporarily delayed. In response, YTL Group, a key client in Malaysia with close ties to Nvidia, has shifted its AI server orders to Wiwynn’s Malaysia facility, a subsidiary of Wistron, for local support.

When asked about the potential order transfer due to the Supermicro incident, Wistron declined to comment on individual clients and products.

In a written statement to the Central News Agency, Supermicro reiterated, “As previously announced, our Malaysia facility is on track to commence production and shipping by the end of 2024. This plan remains unchanged.”

Supermicro further noted, “We continue to secure substantial business in Malaysia within a highly competitive market, meeting all approved orders on schedule, and customer demand remains strong.”

Founded in 1993 by Taiwanese CEO Charles Liang, Supermicro has benefited from the AI boom, posting exponential growth in revenue each of the past three quarters. However, it missed its financial reporting deadline amid media reports in late September alleging accounting issues, now under investigation by the U.S. Department of Justice.

Read more

• [News] Supermicro’s Financial Crisis Shifts Malaysia’s YTL Group Order to Taiwanese Firm

(Photo credit: Supermicro)

As two representative materials of wide-bandgap semiconductors, gallium nitride (GaN) currently thrives in the fast charging sector of consumer electronics, while silicon carbide (SiC) is gradually gaining ground in the application scenarios of new energy vehicles. Simultaneously, both are expanding into broader application boundaries. The strong emergence of AI has created new incremental markets for silicon carbide and gallium nitride.

Against this backdrop, innovative products such as Infineon’s CoolSiC™ MOSFET 400V series and Navitas’ new 4.5kW server power supply solutions have recently been launched, driving the increasing popularity of silicon carbide/gallium nitride in AI server power supplies (PSU).

Silicon Carbide/Gallium Nitride: The Breakthrough for PSU Upgrades

In recent years, the hot application of generative AI and the explosive growth of AI chip computing power have led to a surge in global data center electricity consumption. To address the high energy consumption crisis brought about by AI, upgrading data center PSUs has become a critical breakthrough.

Currently, PSUs for global AI and hyperscale computing data centers come in three form factors: CRPS185, CRPS265, and OCP of the Open Compute Project. The heights and widths of these power supplies are the same, differing only in length. Each CRPS185 power supply has fixed dimensions of 185mm x 73.5mm x 40mm. Thus, given that the physical size cannot change, increasing power demands from AI servers necessitate improvements in power density.

Power density enhancements can be achieved by increasing switching frequency. Since silicon-based products have reached their physical performance limits, the high switching frequency characteristics of GaN devices make them more suitable for high-density CRPS applications. In contrast, SiC devices, compared to silicon-based products, can operate at higher temperatures and voltages, achieving more efficient power conversion and reducing energy loss.

In practical applications, Navitas has significantly improved power density and efficiency through a reference design for CRPS server power supplies that combines SiC power devices and GaN power chips. In July this year, Navitas released an innovative CRPS185 4.5kW AI data center server power supply solution based on its GaNSafe™ high-power GaN power chips and the third-generation fast SiC power devices from GeneSiC™, reportedly achieving a leading power density of 137W/in³ and efficiency of over 97% in the AI data center PSU industry.

The Battle for AI Server Power Supplies Among Leading Silicon Carbide/Gallium Nitride Power Device Manufacturers

A competition in AI server power supplies among SiC/GaN power device manufacturers is already underway. Apart from Navitas, players such as Infineon, ON Semiconductor, EPC, Texas Instruments, GaNext, Corenergy, and Danxitech have entered the fray.

Infineon

In June of this year, Infineon launched a new CoolSiC™ MOSFET 400V series specially developed for AC/DC stages of AI servers. This new series has lower conduction and switching losses compared to existing 650V SiC and Si MOSFETs, increasing the power density of AI server power supplies to over 100W/in³ and achieving an efficiency of 99.5%, which is 0.3 percentage points higher than solutions using 650V SiC MOSFETs.

For the 54V output platform of AI servers, Infineon has developed a 3.3kW PSU dedicated demo board, utilizing Infineon’s CoolGaN™, CoolSiC™, and CoolMOS™ designs to achieve a benchmark overall efficiency of 97.5% and a power density of 96W/in³, meeting the high-power demands of data center PSUs.

At the PCIM Asia 2024 exhibition, an Infineon spokesperson emphasized that reducing data center energy consumption is a pressing need for AI industry development and a primary focus for Infineon. Infineon’s new generation of SiC MOSFET trench gate technology features lower on-resistance, higher switching efficiency, and better reliability, offering significant performance improvements to meet the needs of AI server power supplies.

ON Semiconductor

In response to the “three highs” challenge of PSU output power, conversion efficiency, and power density, ON Semiconductor has launched the latest generation of T10 PowerTrench® series and EliteSiC 650V MOSFET combination, providing a comprehensive solution for data center applications. This solution offers higher efficiency and better thermal performance in a smaller package size.

Among these products, the EliteSiC 650V SiC M3S MOSFET is designed to address data center efficiency challenges, meeting the Open Rack V3 (ORV3) PSU’s peak efficiency requirement of up to 97.5%. The T10 PowerTrench® series enhances thermal performance through optimized packaging technology, addressing the need for high power conversion efficiency and miniaturization in data centers.

EPC

At the PCIM exhibition, EPC showcased samples of humanoid robots using GaN and an autonomous mini-vehicle incorporating EPC’s GaN power devices in its onboard LiDAR components.

In addition to scenarios like humanoid robots and LiDAR, AI data center PSUs are also a key focus area for EPC. In communications with TrendForce Compound Semiconductor, Zhang Shengke, Vice President of EPC’s Reliability Division, stated that EPC’s low-voltage GaN devices can meet all 48V-to-12V server power converter component needs.

TI

As early as 2021, TI had reached a collaboration with server power supply provider Delta to develop efficient, high-power enterprise PSUs for data centers based on TI’s GaN technology and C2000™ MCU real-time control solutions.

TI has invested ten years into GaN technology and C2000™ MCU real-time control solutions. Through collaboration with Delta, TI can utilize innovative semiconductor manufacturing processes to produce silicon-based GaN and integrated circuits, helping companies like Delta create differentiated applications and power data centers worldwide more efficiently.

To lead the industry’s development, both Navitas and Infineon announced their AI data center power supply technology roadmaps, demonstrating their commitment to continuously strengthen their presence in this field.

To meet the exponential power demand increase of AI data centers, Navitas is continuously developing new server power supply platforms, aiming to rapidly elevate power levels from 3kW to 10kW, with a launch expected in Q4 2024.

In addition to its existing 3kW and 3.3kW PSUs, Infineon will soon launch new 8kW and 12kW PSUs to further enhance AI data center efficiency. With the 12kW reference board, Infineon is poised to release the world’s first AI data center PSU to achieve this performance level.

On the domestic front, GaNext has partnered with prestigious universities to jointly develop and mass-produce the industry’s first 3.5kW CRPS fanless server power supply, achieving a high efficiency of 97.6% through two-phase interleaved totem-pole PFC and LLC, with a power density reaching 73.6W/Inch³. Several relevant products from Nengsi Technology have been sampled by leading domestic enterprises and have successfully completed the related reliability tests. Danxitech’s related products have been sampled to leading domestic companies and have completed reliability tests. Corenergy ‘s 1200V GaN products have been sampled by a known server power supplier and are undergoing reliability evaluation. Various major manufacturers are jointly advancing the industrialization of GaN in AI data center PSU applications.

Summing up, power density and efficiency in AI server power supplies continue to improve, focusing on two primary performance indicators for SiC/GaN power device manufacturers. This trend is likely to spur the production of various high-performance products, attracting more SiC/GaN players to enter the market.

Conclusion

The data center power supply sector has become a key focus of leading silicon carbide/gallium nitride manufacturers in recent years. Some manufacturers’ SiC/GaN products have already made progress in the data center power supply market, with the strong rise of AI further driving the development of this market.

As AI technologies evolve and computational power demand continues to grow, the energy efficiency and power density requirements of data centers will become increasingly stringent, imposing higher performance demands on SiC/GaN power device products from various manufacturers.

The continuous development of the SiC and GaN industry, coupled with the thriving AI ecosystem, will result in more intersections between SiC, GaN, and AI. The competition for AI server power supplies among major manufacturers will become increasingly intense.

Though previous rumors suggested that there might be a downward revision of its HBM capacity target due to delayed progress on 12-Hi HBM3e, Samsung’s officials have stated that it will enhance its semiconductor packaging facilities in South Chungcheong Province to increase the production of HBM, according to a report by the Korea Herald.

Samsung aims to complete the new facilities by December 2027, which will feature advanced packaging lines for HBM chips, the Korea Herald report indicates.

However, it is worth noting that the capacity is not built from scratch. According to the report, Samsung will repurpose an underutilized liquid crystal display plant, previously owned by Samsung Display, into a semiconductor fabrication facility. The plant is said to be located in Cheonan, approximately 85 kilometers south of Seoul.

According to the report, Samsung anticipates that the upgraded facilities in Cheonan will help the company regain its competitive edge in the global semiconductor market.

The current HBM leader, SK hynix, is reportedly investing in advanced chip packaging as well, as it aims to capture more demand for HBM in the AI boom.

According to a previous report by Bloomberg, Lee Kang-Wook, currently leading SK Hynix’s packaging research and development, stated that the company is investing over USD 1 billion in South Korea to expand and enhance the final steps of its chip manufacturing process.

On the other hand, Micron stated last year that in response to the growing demand in the AI market, it will continue to invest in advanced processes and packaging technologies to produce HBM products. Micron Taiwan is reportedly the only Micron facility globally with advanced packaging capabilities.

This August, Taiwanese panel maker AUO announced that it will sell three idled manufacturing facilities in Tainan, Southern Taiwan as Micron emerged as the buyer. Micron is also mulling HBM expansion in Malaysia and the U.S.

Read more

• [News] SK hynix and Samsung Reportedly Step up Focus on HBM4 and CXL amid Rising Chinese Competition
• [News] Samsung Reportedly Starts Developing Custom HBM4 for Big CSPs, Eyeing Mass Production by 2025

(Photo credit: Samsung)

As global economic growth slows, many central banks around the world have begun to cut interest rates in an effort to reignite economic expansion. One of the main reasons driving this move is that the year-on-year inflation rate in most countries has gradually declined from its post-pandemic peak to their targeted “2%” level.

In theory, moderating price growth should bolster consumer optimism, as it suggests that inflation’s erosion of wage gains is easing. Yet why do media reports still frequently highlight consumers struggling under the weight of high living costs?

The answer lies in the fact that the inflation rate measures year-over-year price changes. While current price growth may have slowed compared to the previous year, it does not change the fact that prices have significantly risen from the pre-pandemic levels.

(Source: BLS, EuroStat, Statistic Bureau of Japan, TrendForce)

Take the United States as an example. The latest CPI and core CPI year-on-year growth rates were 2.4% and 3.3%, respectively, marking their slowest pace in nearly three years. Meanwhile, the Federal Reserve’s preferred inflation gauge, core PCE, stood at 2.7%, down significantly from its peak of 5.6% in February 2022, reflecting a 2.9 percentage point decline.

Read more at Datatrack

However, consumer perceptions often remain anchored to periods of lower prices, making it difficult to truly appreciate the recent moderation in price growth. From the perspective of consumers, prices for goods are still higher than they were a few years ago, even if the pace of increase has slowed.

According to average price data from the U.S. Bureau of Labor Statistics, most goods have seen price increases of around 20-40% over the past few years. Importantly, these elevated prices are unlikely to revert unless deflation occurs—something the Federal Reserve is unlikely to allow.

(Source: BLS)

Additionally, while wage growth has outpaced inflation since mid-2023, overall wage increases since the pandemic have remained below the cumulative rise in inflation. This has been a key factor in the decline in consumer confidence and the frustration with persistently high inflation in recent years.

(Source: BLS, Fred, TrendForce)

According to a report from Liberty Times, citing the Reuters, on Monday (11th), South Korea’s ruling party proposed a special semiconductor bill to provide subsidies for chip manufacturers and remove the national cap on working hours, addressing potential risks stemming from Trump’s threats of tariffs and chip-related measures.

The bill would allow some employees involved in R&D to work longer hours, exempting them from the labor law’s 52-hour weekly work limit, as noted by the report.

The report indicated that as Asia’s fourth-largest economy, South Korea is highly dependent on trade, with the semiconductor industry playing a critical role. According to the Reuters report, chips accounted for 16% of South Korea’s total exports last year.

One of the bill’s sponsors, lawmaker Lee Chul-gyu, stated that China, Japan, Taiwan, and the U.S. are all subsidizing manufacturers amid the semiconductor trade war between the U.S. and China, and this bill will help South Korean companies face these challenges, as indicated by the report from the Reuters.

Last week, according to the report, South Korean President Yoon Suk Yeol cautioned about the risks arising from Trump’s threat to impose steep tariffs on Chinese imports, which could lead Chinese competitors to lower export prices and undermine Korean chip companies in international markets.

The report noted that the ruling party’s bill comes as chipmakers like Samsung Electronics face increasing competition from rivals in Taiwan, China, and other countries.

On the other hand, the report pointed out that the bill proposed by the ruling party still requires approval from the main opposition party to pass.

Samsung’s labor union has also voiced opposition, arguing that the company is using the law as an excuse for its “management failure,” as the report noted.

The report pointed out that, last month, Samsung issued an apology for its disappointing profits, acknowledging that it had fallen behind competitors TSMC and SK Hynix in capitalizing on the surging demand for AI chips.

Read more

• [News] South Korea’s Labor Regulations May Weaken its Semiconductor Competitiveness
• [News] Samsung Reportedly Joins TSMC in Halting 7nm and Below AI Chip Supplies to China

(Photo credit: Samsung)

According to a Reuters report, the Japanese government has proposed a USD 65.1 billion (JPY 10 trillion) plan to boost the domestic chip and AI industry through subsidies and other financial support.

The plan aims to provide support totaling USD 65 billion (JPY 10 trillion) by fiscal 2030, with the goal of strengthening Japan’s control over its chip supply chain in response to potential impacts from U.S.-China trade tensions, as noted in the report.

The report indicated that Japan’s government is set to submit this plan to the next parliamentary session, and the draft includes financial support for mass production of next-generation chips, specifically targeting Rapidus and other AI chip suppliers. The report highlighted that, the government anticipates an economic impact of approximately JPY 160 trillion according to the draft.

Rapidus is scheduled to begin mass production of cutting-edge chips in Hokkaido starting in 2027, in collaboration with IBM and the Belgium-based research organization Imec, as the report pointed out.

According to the report, Japanese Prime Minister Shigeru Ishiba stated that the government would not issue deficit-covering bonds to fund the chip industry support plan. However, detailed information on how the plan will be financed has not been disclosed.

Last year, the Japanese government announced that it would allocate approximately JPY 2 trillion to boost its chip industry, as the report noted.

The latest plan is part of a comprehensive economic package anticipated for Cabinet approval on November 22, calling for a total investment of JPY 50 trillion in the chip industry from both public and private sectors over the next 10 years, according to the report from the Reuters.

Read more

• [News] Japanese Chipmaker Rapidus to Receive EUV Machine by Year-end, Boosting Hokkaido’s Economy
• [News] Sony and Kioxia Reportedly Plan Additional Investments in Chip Foundry Rapidus

(Photo credit: Rapidus)

With Trump’s inauguration in January, “Made in America” is expected to dominate, driving an urgent push for semiconductor packaging and equipment companies to relocate to North America. Coupled with TSMC’s Arizona plant set to start mass production early next year, the Commercial Times reports that industry insiders foresee a surge in investments across the North American semiconductor supply chain.

TSMC’s Arizona fab is in the final stages of preparing for 4nm production, with a projected monthly capacity of 20,000 to 30,000 wafers. The company previously signed an MOU with Amkor Technology, a partnership widely seen as supporting advanced and back-end packaging efforts.

On November 8, however, leading packaging and testing company ASE also announced plans to set up a facility in Mexico, aiming to offer advanced packaging services for TSMC’s U.S.-produced chips.

Industry sources cited by Commercial Times speculate that ASE’s Mexico plant, once complete, could compete with Amkor for TSMC’s packaging and testing orders from the Arizona fab. Following the packaging process, the chips could be delivered directly to U.S.-based OEM/ODM partners, including Foxconn, Wistron, and Inventec, for final product assembly, completing the “Made in America” manufacturing chain.

The report also highlights TSMC’s stronghold on advanced packaging technologies such as 3D Fabric and SoIC, required for 2nm production, as well as SoW (system-on-wafer) technology. To meet customer demands, TSMC may need to establish in-house advanced packaging capabilities in the U.S. for even more advanced processes.

TSMC’s advanced packaging line is already highly automated, with optimized production flows reduced from over 300 steps to just over 200, and its gross margin is approaching the company average. The Commercial Times quotes industry experts who suggest that setting up advanced packaging capacity in the U.S. should be straightforward for TSMC, given its extensive experience in wafer fab construction, making it a matter of time.

(Photo credit: TSMC)

The U.S. has tightened restrictions on foundries supplying 7nm and below chips to Chinese clients. According to a report from Economic Daily News, following TSMC’s suspension of services to restricted Chinese clients, rumors suggest that South Korea’s Samsung is also affected by U.S. restrictions, preventing it from offering such foundry services to Chinese firms.

Samsung’s foundry division has reportedly notified Chinese clients about the restrictions, according to the report. However, Samsung has declined to comment on these rumors.

As for Intel, since it is closely aligned with the U.S., the company is also expected to comply with U.S. regulations, as the report mentioned. This would lead to a comprehensive blockade of China’s effort to develop advanced AI chips, signaling a new chapter in the U.S.-China semiconductor confrontation, potentially reshaping the global semiconductor landscape.

The report pointed out that currently, only three companies—TSMC, Samsung, and Intel—are capable of providing foundry services below the 7nm process. China’s leading foundry, SMIC, claims to have 7nm production capability, but it lacks the necessary economies of scale and efficiency.

Citing industry sources, the report suggests that Alibaba’s AI chip subsidiary, T-Head, could be the most heavily impacted by the restrictions, indicating that, following Huawei, Alibaba has now also come under scrutiny by the White House. Along with T-head, other Chinese AI chip companies, such as Bitmain and Cambricon, are also likely to be impacted.

Fully owned by Alibaba, T-Head has made rapid progress in next-generation chip development, with its “Yitian” series reaching sub-5nm technology, as the report mentioned. The company claims that its technology supports applications across diverse fields, including AI in automotive, gaming, and scientific research.

The report indicated that as China’s path to self-developed AI chips encounters setbacks, it may be forced to rely on downgraded versions of AI chips from U.S. companies like NVIDIA and AMD to comply with American restrictions. This development, as the report noted, could undermine Beijing’s hopes of using domestically developed AI chips to circumvent U.S. limitations.

(Photo credit: Samsung)

Read more

• [News] Four Key Takeaways on TSMC’s Reported Halt of 7nm and Below Chip Supplies to China
• [News] TSMC Reportedly Halts Shipments to Chinese Firm Sophgo After Chip Found in Huawei Processor

As Samsung has implied earlier that it plans to collaborate with TSMC on HBM4, the memory giant seems to take a key step forward. A report by South Korean media outlet Maeli Business Newspaper discloses that the memory giant has already begun developing “Custom HBM4,” a next-gen high-bandwidth memory tailored specifically for Microsoft and Meta.

Samsung aims to begin mass production of HBM4 promptly upon completing development by the end of 2025, the report suggests.

Industrial sources cited by the report state that Samsung is establishing a dedicated production line for HBM4, which is now in the “pilot production” phase, where small-scale, trial manufacturing takes place ahead of mass production.

Citing sources familiar with the situation, the report further notes that Samsung is actively working on HBM4 designed for Microsoft and Meta. Both tech heavyweights have their own AI chips, namely, Microsoft’s Maia 100 and Meta’s Artemis.

As major tech companies make strides in scaling up their own AI data centers, there is a strong demand to cut costs associated with chip purchases. Therefore, many design and utilize their own AI accelerators while also buying AI chips from NVIDIA and AMD, according to Maeli.

Samsung, with its memory division and an LSI division capable of designing computational chips, is ideally positioned as a top partner for these major tech companies, according to Maeli.

Though the specifics of the HBM4 product that Samsung will supply to these companies remain undisclosed, Samsung did reveal the general specifications in February, according to Maeli.

Its data transmission speed, or bandwidth, reportedly reaches 2 terabytes per second (TB), marking a 66% increase over HBM3E, while its capacity has risen 33% to 48 gigabytes (GB), up from 36GB.

The report explains that unlike previous memory products, HBM4 is also referred to as “Computer-in-Memory (CIM)” due to its advanced capabilities, which go beyond standard memory functions to include specialized operations aligned with customer requirements.

Up to HBM3, the focus is said to be mainly on managing heat generation and maximizing speed, according to a semiconductor industrial official cited by the report. With HBM4, however, the emphasis will shift toward integrating AI computation capabilities (NPU) and specific data processing functions (IP) to meet evolving needs, the report says.

Read more

• [News] SK hynix and Samsung Reportedly Step up Focus on HBM4 and CXL amid Rising Chinese Competition
• [News] SK hynix Reportedly to Tape-out HBM4 in October, Paving the Way for NVIDIA’s Rubin

(Photo credit: Samsung)

Currently, the global semiconductor industry is entering a new period of transformation. With the rapid development of AI, big data, cloud computing, and other technologies, the demand for high-performance computing chips, optical communication chips, and advanced packaging has surged, and recent reports suggest that prices for these types of chips are increasing.

Advanced Process and Advanced Packaging Products May See Price Increases

According to a recent report by Morgan Stanley, TSMC is considering raising prices for its 3nm process and CoWoS advanced packaging technology in response to soaring market demand. TSMC plans to implement these price increases in 2025, with the cost of its 3nm process potentially rising by up to 5%.

Industry analysts point out that on the demand side, major AI chip manufacturers such as NVIDIA and AMD heavily rely on TSMC’s 3nm process, and the explosive growth in AI technology has driven continuous demand for these chips, contributing to the price increase.

On the supply side, the high research and production costs associated with advanced process technology—including equipment investment, material costs, and R&D personnel—add significant pressure to the supply chain. Multiple factors have led to a tight supply of such chips, further driving up prices.

Additionally, TSMC’s 5nm and 4nm process quotes have increased more than previously anticipated by 4%, with some price hikes reaching as much as 10%.

Reports indicate that TSMC also plans to raise prices for its CoWoS advanced packaging technology, with potential increases between 10% and 20%. High demand for CoWoS from major companies like NVIDIA, AMD, Microsoft, Amazon, and Google has resulted in a shortage of CoWoS packaging capacity, which has driven up prices.

According to TrendForce research, NVIDIA is the primary driver of demand for CoWoS, and with the upcoming launch of its Blackwell series, demand for CoWoS is expected to increase by more than 10 percentage points annually by 2025.

Optical Communication Chip Sector Begins Price Increases

Demand for high-speed, high-bandwidth, and low-latency optical communication is rising, particularly in data centers, enterprise networks, and telecommunications, driving demand in the optical communication chip market. Recently, media reports revealed that Marvell, a major optical communication chip manufacturer, has issued a price increase notice, with its entire product line set to see price hikes starting January 1, 2025. According to TrendForce, Marvell ranked sixth in the global IC design market in 2023.

Industry forecasts predict that, driven by ongoing advances in optical communication technology and expanding applications, the global optical communication chip market will grow rapidly in the coming years.

The development and application of technologies such as silicon photonics, optoelectronic hybrid integration, and high-performance photonic chip materials are expected to bring new growth points and opportunities to the optical communication chip market.

(Photo credit: Marvell)