The COVID-19 pandemic heavily impacted the global auto market and in turn damaged the automotive LED industry in 1H20, according to TrendForce’s latest investigations. In 2H20, however, the gradual recovery of vehicle sales as well as the development of NEVs provided some upward momentum for the automotive LED industry, whose revenue for the year reached US$2.572 billion, a 3.7% decline YoY. Automotive LED revenue for 2021 is projected to reach $2.926 billion, a 13.7% growth YoY, thanks to the increasing demand for automotive headlights and display panels. As automakers continue to incorporate LED lighting solutions into new car models, the penetration rate of automotive LED will continue to undergo a corresponding increase as well.

TrendForce analyst Joanne Wu indicates that, in the automotive LED player revenue ranking of 2020, OSRAM Opto Semiconductors, Nichia, and Lumileds remained the top three largest automotive LED suppliers, respectively, with a combined market share of 71.9%. In particular, European and American automakers favored OSRAM’s solutions for their high-end vehicle models and NEVs due to the high quality of OSRAM products. Adoption by these automakers subsequently became the main revenue driver of OSRAM’s automotive LED business.

On the other hand, the pandemic caused Japanese automakers to suspend their operations and therefore had a direct impact on the revenues and market shares of Japanese LED suppliers, such as Nichia and Stanley, in 2020. Nichia and Stanley saw their revenues decline by 9.8% YoY and 7% YoY, respectively, and were the two suppliers among the top 10 last year to have shown relatively noticeable declines. Seoul Semiconductor’s nPola and Wicop LED products were adopted by Chinese automakers, including CCAG, SAIC-GM, and NIO, due to these products’ high brightness and compact sizes. Seoul Semiconductor’s market share reached 5.1% in 2020. Finally, not only did other suppliers, including Samsung LED and CREE, deliver consistent performances in the automotive aftermarket (AM) and performance market (PM) segments, but they also gradually began entering the automotive Original Equipment Manufacturer (OEM) lighting market. Samsung LED and CREE each took seventh and ninth place on the 2020 ranking with a 2.8% and 1.1% market share, respectively.

On the whole, TrendForce finds that automotive demand has been recovering since 4Q20. Accordingly, LED suppliers indicate that their booking orders appear bullish throughout 2021, meaning most LED suppliers now need to extend their product lead times in response. At the same time, LED players indicated that double booking might happen in the near future. Thereby, they will make decisions in light of the actual booking order quantity to see the possibility of increasing prices.

(Cover image source: OSARM)

For more information on reports and market data from TrendForce’s Department of Optoelectronics Research, please click here, or email Ms. Grace Li from the Sales Department at graceli@trendforce.com

Owing to demand generated by the persistent stay-at-home economy last year and from the emerging markets in certain developing countries, global TV shipment for 2021 is expected to reach 223 million units, a 3.1% growth YoY, according to TrendForce’s latest investigations. The delay of UEFA Euro 2020 and the Tokyo Olympics until this summer will likely also play a role in driving up TV demand, regardless of whether live attendance will be allowed at the events. However, prices have increased repeatedly and considerably for not only IC components (used in TV set assembly), which are in shortage due to tight foundry capacities, but also TV panels. The price hike of TV panels has persisted since last June, with 32-inch panels, which are indicative of the rest of the TV panel market, reaching a massive 134% price hike for the period.

TrendForce’s investigations also show that the increase in panel prices has made it difficult for white-label manufacturers and tier 2/3 brands, which have traditionally relied on aggressive pricing to achieve their sales performances, to procure sufficient panels. Case in point, TV shipment from these companies has been gradually declining since last year. Conversely, suppliers have been giving major TV brands top priority ahead of the aforementioned companies to procure both panels and components because major TV brands generally place orders regularly and in large quantities. For the first time ever, the combined market shares of the top five brands, which are Samsung, LG, TCL, Hisense, and Xiaomi, surpassed 60% last year. This figure is expected to further increase to 62% in 2021, representing the fact that the TV market is progressively becoming an oligopoly.

As brands begin to favor large-sized products, 60-inch and larger TVs are expected to account for 17.7% of total TV shipment for the first time ever

With regards to various TV sizes, 32-inch panels have more than doubled in price since the start of the upturn last June. In response, TV brands have been transitioning their product lines to TVs that are at least 55 inches in size. More specifically, 55-inch TVs and ultra-large-sized TVs (60-inch and above) will account for 20% and 17.7% of the total TV shipment this year, respectively. Whereas the 20% shipment share of 55-inch TVs remained the same as last year, the 17.7% shipment share of ultra-large-sized TVs is 3.3% higher than last year’s figure. With regards to the annual shipment of ultra-large-sized TVs, 2021 marks a year of considerable growth compared to previous years, which generally saw YoY increases of 1-2%. This growth reflects the necessity for TV manufacturers to quickly leverage the consumer demand for large-sized TVs in order to maintain a stable growth in the industry, given the substantial price hike of TV panels.

As the difference between OLED and LCD panel prices narrows, TV brands are compelled to accelerate their OLED TV strategies

In response to the massive price hike of LCD panels, TV brands have begun to slightly raise the retail prices of TVs across various segments in order to keep up their bottom lines. However, if brands were to at once completely offload the increase in panel prices to the retail end, consumer demand would plummet as a result. A slow and gradual price hike is therefore expected to take place instead. Incidentally, it should be pointed out that the price hike of TV panels would be unlikely to stop in the short run even if the current panel shortage were alleviated in the future. As such, TV brands are expected to have limited room for profit growth in 2Q21.

The OLED panel market, on the other hand, has taken an opposite turn compared to the LCD market. For instance, prices of 55-inch UHD OLED panels were four times the prices of equivalent LCD offerings at the start of 2020, 2.9 times at the end of 2020, and 2.2 times in 1Q21, while prices of LCD panels underwent monthly increases. In addition to the narrowing gap between OLED and LCD panel prices, the panel industry’s production capacity for OLED panels saw a major boost thanks to the capacity expansion of LGD’s Gen 8.5 fab in Guangzhou. TrendForce expects OLED TV shipment for 2021 to reach 6.76 million units, a staggering 72% increase YoY, as OLED offerings become the top strategic priorities of TV brands in the high-end TV market this year.

For more information on reports and market data from TrendForce’s Department of Display Research, please click here, or email Ms. Vivie Liu from the Sales Department at vivieliu@trendforce.com

A fire broke out at the 12-inch wafer production line of Renesas’s Naka Factory on March 19 due to an overcurrent in the plating equipment. Renesas said that the fire burned about 5% of the total area of the first floor. The Naka fab mainly manufactures MCUs and SoCs for automotive, industrial, and IoT-related applications. While Renesas officially aims to get the fab back to full operation within one month, TrendForce expects the immediate task of restoring the cleanroom and installing new equipment systems to take much longer than that. The repair of the production line will have to proceed meticulously so as to avoid the risks of manufacturing-related problems in the mass production of automotive chips later on. Three months is TrendForce’s conservative estimate for the fab to regain its former level of wafer-start capacity, meaning the tight supply of automotive MCUs will be further exacerbated going forward.

The Naka incident is not expected to result in additional orders for other foundries, given the current tight wafer-start capacity across the foundry industry

TrendForce indicates that the 12-inch Naka fab’s process technologies likely range from the 90nm node to the 40nm node. With regards to Renesas’s production lines for automotive chips, TrendForce expects the fire to impair the fab’s wafer-start capacities for products including automotive PMICs, certain V850 automotive MCUs, and first-generation R-Car SoCs. Other foundries, in particular TSMC, are able to support some of Renesas’s production, since 2/3 of their technologies are interoperable. However, it is exceedingly difficult for other foundries to allocate spare wafer-start capacities to make up for Renesas’s shortfall due to the existing wafer-start capacity crunch across the foundry industry.

Ranked third among automotive semiconductor suppliers in 2020, Renesas is also currently one of the top five largest automotive MCU suppliers at the moment. Other automotive MCU suppliers include STMicroelectronics, Infineon, NXP, TI, and Microchip. Although most of STMicroelectronics’ automotive MCUs are manufactured in-house, TrendForce believes that the Naka fire will not result in additional orders for Renesas’s competitors, including STMicroelectronics, since automotive semiconductors are currently in extreme shortage.

For more information on reports and market data from TrendForce’s Department of Semiconductor Research, please click here, or email Ms. Latte Chung from the Sales Department at lattechung@trendforce.com

The x86 architecture remained the mainstream server architecture at the end of 2020, according to TrendForce’s latest investigations. In the x86 server segment, Intel took the lead with a 92% market share thanks to the wide-ranging positioning of its solutions. On the other hand, AMD saw its market share rise to nearly 8% in 4Q20, which represents a 3% growth compared to 4Q19. Other server solutions, based on non-x86 architectures, comprised an insignificant portion of the market. TrendForce projects total server shipment to increase by 21% QoQ in 2Q21 owing to the release of Intel’s new Ice Lake platform.

Notably, edge computing, which involves low-latency data processing, has risen to the forefront of server applications in response to the rise of AI and 5G networks. Ideal use cases of edge computing include autonomous driving, IIoT, and other proprietary commercial applications, with relevant vendors already scrambling to deliver solutions aimed at these use cases. With regards to architecture, ARMv8-based solutions are the most suitable for edge computing applications in terms of both cost and power consumption. With the ongoing 5G commercialization, ARMv8 solutions have been occupying a growing share within the total shipment of all ARM solutions, and North American data centers have been the most aggressive in adopting these solutions. TrendForce expects ARMv8 solutions to occupy a respectable share of the low power consumption edge computing market in 2023-2025.

With the upcoming ramp-up in production, Intel’s new Ice Lake CPUs will likely account for 40% of Intel CPU shipment in 4Q21

As previously mentioned, Intel’s x86 CPUs are the mainstream among server platforms. With regards to Intel’s server roadmap, the company has started shipping a small batch of Whitley Ice Lake CPUs to its clients in the data center segment in 1Q21, while bulk shipment to branded server manufacturers is expected to take place in 2Q21. At the same time, Intel expects Whitley Ice Lake CPUs to account for about 40% of Intel’s CPU shipment in 4Q21, although adoption by buyers remains to be seen.

In terms of specifications, the Whitley platform allows Ice Lake CPUs to be compatible with the fastest DDR4 standards and therefore provides a major improvement in both data transfer rate (MT/s) and maximum capacity of DRAM supported per CPU. The Whitley platform will effectively increase the average DRAM storage capacity of servers, leading to faster and more advanced virtualization applications while improving data streaming performances in data centers. On the whole, not only does the Whitley platform increase the Ice Lake CPU’s cloud computing capabilities, but it also allows Intel to catch up to AMD’s Rome platform, in turn further strengthening Intel’s leadership position in the server market. As such, buyers will likely adopt the Whitley platform given the increased server demand generated by the post-pandemic new normal. Incidentally, although AMD’s mass-produced server solutions are slightly superior to competing offerings from Intel in terms of specs and pricing, AMD will not make significant adjustments in its upcoming Milan platform. AMD will therefore unlikely be a factor that propels the overall bit demand for server DRAM.

In the competition among server CPUs with respect to the data transfer rate, the Whitley Ice Lake is Intel’s first server processor platform that supports PCIe Gen 4. Although AMD was already mass producing server CPUs featuring PCIe Gen 4 support in 2020, Intel will likely be more effective in leveraging this advantage across the application ecosystem. The reason is that Intel is the long-time leader in server CPUs and controls the lion’s share of the market. Furthermore, Intel aims to synergize its latest server CPUs with the second generation of its Optane SSDs. Working together, they are expected to significantly enhance the computing performance of servers for the applications that will become mainstream in the future (e.g., AI and machine learning). TrendForce believes that the penetration rate of PCIe Gen 4 will rise rapidly in 2H21 as Intel and AMD step up shipments of CPUs that support this interface.

Regarding the implementation of the support for DDR5 and PCIe Gen 5, both Intel and AMD have already begun sending samples of related products to their clients in 1Q21. Intel plans to begin mass production for server CPUs belonging to the Eagle Stream platform at the start of 2022. Since the Eagle Stream is designed to support PCIe Gen 5, it will provide another significant boost to the data transfer rate. As for AMD, it plans to commence mass production for server CPUs based on the Genoa platform in 2Q22.

For more information on reports and market data from TrendForce’s Department of Semiconductor Research, please click here, or email Ms. Latte Chung from the Sales Department at lattechung@trendforce.com

With Samsung, YMTC, SK Hynix, and Intel leading the charge, NAND Flash suppliers will maintain an aggressive effort to expand their production capacities throughout 2Q21, during which NAND Flash bit output will likely increase by nearly 10% QoQ, according to TrendForce’s latest investigations. On the other hand, orders from PC OEMs and Chinese smartphone brands since 1Q21, as well as recovering procurement activities from clients in the data center segment during 2Q21, will generate upward momentum propelling NAND Flash bit demand. Furthermore, buyers are actively stocking up on finished products, such as SSDs and eMMC, due to persistently limited NAND Flash controller supply. TrendForce therefore expects NAND Flash contract prices to increase by an average of 3-8% QoQ in 2Q21 after experiencing a 5-10% decline QoQ in 1Q21. In particular, as Samsung’s Line S2 fab in Austin has yet to resume full operation after the Texas winter storm, the supply of NAND Flash controllers going forward may be at risk, and Samsung’s ability to manufacture client SSDs will be further constrained as a result. In light of these factors, TrendForce is not ruling out the possibility that NAND Flash contract prices may increase by even more than current forecasts.

Contract prices of both client SSDs and enterprise SSDs are projected to rise due to delayed resumption at Samsung’s Line S2 fab

With regards to client SSDs, the persistent stay-at-home economy generated by the COVID-19 pandemic will likely result in strong demand for notebook computers in 2H21, while PC OEMs have raised their client SSD inventories as they manufacture more notebooks to meet demand. Given the high volumes of client SSD orders from PC OEMs, inventory level of NAND Flash suppliers is therefore likely to remain healthy. However, the shortage of NAND Flash controllers has yet to be resolved. Suspended operations at the Line S2 fab disrupted Samsung’s production of NAND Flash controllers, meaning some client SSD orders will not be fulfilled in 2Q21. Hence, the tight supply of finished products (i.e., client SSDs) will be further exacerbated. As such, client SSD contract prices are projected to increase by 3-8% in 2Q21.

With regards to enterprise SSDs, demand is expected to rebound from rock bottom in 2Q21, primarily because clients in the data center segment will ramp up their procurement activities after undergoing a period of inventory adjustment. In addition, demand for IT equipment from the governmental, healthcare, and financial services sectors will also gradually emerge. Other factors contributing to enterprise SSD demand include bids from Chinese telecom operators and increased IT equipment purchases from small and medium businesses globally. On the other hand, NAND Flash suppliers are no longer under pressure to destock via low prices, since their inventory levels have improved thanks to high demand from notebook manufacturers and smartphone brands. As the overall demand for NAND Flash rises, enterprise SSD contract prices are in turn expected to stabilize and experience a 0-5% growth QoQ in 2Q21.

High demand for Chromebooks will provide upward momentum for eMMC quotes, while contract prices of UFS are projected to undergo the lowest growths among NAND Flash products

eMMC contract prices will likely remain, for the most part, higher than expected despite the cyclical downturn in 1H21. In particular, strong demand from Chromebook manufacturers will provide upward momentum for eMMC quotes. Likewise, under the influence of NAND Flash controller shortage, eMMC buyers such as consumer electronics manufacturers will expand their procurement activities in order to build up their inventories. As a result, the overall eMMC demand will gradually ramp up in 2Q21. Conversely, the supply of eMMC controllers is still in shortage due to the fully loaded capacities across the foundry industry. Also, eMMC products under 32GB exclusively feature 2D NAND or 64L 3D NAND. Because production capacities allocated for these types of NAND Flash memories have been either reassigned to other 3D NAND products or scaled down, the oversupply of eMMC has been alleviated, and the long-term price drop of eMMC has subsequently come to a halt. In the short term, the shortage of controller ICs will result in a shortage of finished eMMC products. eMMC contract prices are therefore projected to increase by 3-8% QoQ in 2Q21.

Demand for UFS, which is primarily used for smartphones, is expected to remain high through 2Q21 because OPPO, Vivo, and Xiaomi have been aggressively procuring UFS since 4Q20, and Huawei spun off its Honor smartphone business unit. Buyers have also been anticipating an upcoming shortage of controller ICs and NAND Flash memory, leading them to build up their UFS inventories and therefore further driving up the overall UFS demand. On the supply side, inventory levels of suppliers have dropped significantly due to smartphone brands’ large-scale procurement activities previously. Although Chinese smartphone brands have yet to ramp up their bit demand, their existing level of demand still remains strong. Furthermore, clients from the data center segment are expected to increase their SSD procurement in 2Q21, and suppliers will maintain an aggressive approach regarding quotes in response. Even so, because smartphones account for the highest bit consumption share among all NAND Flash applications, NAND Flash suppliers are unlikely to significantly adjust their UFS quotes. As such, UFS contract prices are expected to increase by 0-5% QoQ in 2Q21, which is relatively lower compared to other NAND Flash products.

NAND Flash wafer contract prices are projected to increase by 5-10% QoQ as NAND Flash suppliers lower their bit shipment to the wafer market due to its lower profit margins

With regards to the NAND Flash wafer market, TrendForce has yet to observe an obvious improvement in the retail sales of end products such as SSDs, memory cards, and USB flash drives. However, as NAND Flash suppliers have been unable to make their scheduled delivery dates to OEMs due to an insufficient supply of controller ICs, module makers may stand to benefit from this and obtain more orders from OEMs, subsequently driving up the demand for NAND Flash wafers within the next one to two quarters, but the actual procurement of NAND Flash wafers will depend on whether the tight supply situation of controller ICs can be alleviated. On the other hand, inventories of NAND Flash suppliers have now fallen to mostly healthy levels thanks to procurement activities from smartphone brands since 4Q20. Suppliers have accordingly lowered their bit shipments to the NAND Flash wafer market (which yields a relatively lower profit margin compared to other product categories), due to the rising demand from notebook manufacturers and the expected recovery of the data center segment in 2Q21. On the whole, given the bullish market of mainstream products, such as smartphones and notebooks, TrendForce expects NAND Flash wafer contract prices to once again increase by 5-10% QoQ in 2Q21.

For more information on reports and market data from TrendForce’s Department of Semiconductor Research, please click here, or email Ms. Latte Chung from the Sales Department at lattechung@trendforce.com