The quarterly total revenue of the NAND Flash industry came to US$14.1 billion in 4Q20, showing a QoQ drop of 2.9%, according to TrendForce’s latest investigations. The total bit shipments of the NAND Flash industry in 4Q20 registered a QoQ increase of nearly 9%. This gain for the most part offset the negative impacts brought about by the QoQ decline of nearly 9% in the overall ASP of NAND Flash products, as well as by the unfavorable exchange rates that impaired some suppliers’ performances. At the same time, clients in the server and data center segments continued their inventory reduction efforts that had begun in 3Q20. Since their procurement drive remained fairly weak, contract prices of NAND Flash products continued their decline as well. However, NAND Flash suppliers were receiving substantial orders from OPPO, Vivo, and Xiaomi. On the whole, the strong demand in the smartphone segment mostly compensated for the weak demand in the server and data center segments. In the notebook computer segment, Chromebook devices were the primary sales driver, but the storage components of most Chromebooks are low-density solutions, meaning related NAND Flash demand is somewhat limited.

Turning to 1Q21 (this quarter), the bit output of the NAND Flash industry continues to grow significantly due to two factors. First, Samsung and YMTC are actively expanding production capacity. Second, all suppliers are eager to migrate to the more advanced stacking process. On the demand side, PC and smartphone brands have kept stocking up on components. However, they have also slightly corrected down their demand as the first quarter is the traditional off-season. As for clients in the server and data center segments, they have yet to restart large-scale procurement even though their inventories have generally returned to a healthy level. As such, during price negotiations, NAND Flash suppliers still expect the oversupply situation in the market to intensify and thereby further drive contract prices of NAND Flash products downward. Therefore, NAND Flash suppliers’ revenues are projected to undergo a QoQ decline in 1Q21.

Samsung

Two factors helped Samsung’s performance in the NAND Flash market during 4Q20. First, Chinese smartphone brands (with the exception of Huawei) continued to aggressively build up their component inventories so as to fight for more market share. Second, PC OEMs released more upside orders than anticipated because of the further demand growth for notebook computers. Samsung’s NAND Flash bit shipments in 4Q20 rose by 7-9% QoQ as a result of the strong procurement momentum in the smartphone and PC segments. However, the ASP of its NAND Flash products dropped by more than 10% QoQ for the same period. Although clients in the data center segment did begin sending out orders for servers and components at that time, their demand was still very limited. As for clients in the enterprise server segment, they were mainly focused on inventory reduction. With the market leaning toward oversupply, Samsung had to lower prices and thereby experienced a revenue decline. Compared with 3Q20, Samsung’s NAND Flash revenue fell by 3.4% to US$4.644 billion.

Regarding production and technology plans, Samsung this year will be the leader among NAND Flash suppliers in raising production capacity. Besides continuing to expand the production capacity of the Xi’an Fab 2, Samsung will also set up a production line for 3D NAND Flash at P2L (in the Pyeongtaek campus). Most of Samsung’s NAND Flash products are still based on the V5 (92L) process at this moment. However, the supplier will significantly raise the output share of NAND Flash from the V6 (128L) process this year. The application of the V6 process technology will expand to more of its offerings for SSDs and UFS solutions.

Kioxia

Following the end of Huawei’s stock-up activities, Kioxia was unable to fully regain the demand for its mobile NAND Flash products through orders from other Chinese smartphone brands in 4Q20. At the same time, the supplier was affected by the weak demand for enterprise SSDs. On the other hand, there were other sources of demand such as notebook computers and game consoles. Consequently, Kioxia’s bit shipments in 4Q20 still registered a small increase. As for the ASP, Kioxia experienced a QoQ decline of 8-10% because of the general oversupply situation. On account of these factors, Kioxia’s NAND Flash revenue slid by 11.4% QoQ to US$2.749 billion for 4Q20.

Regarding production and technology plans, Kioxia will gradually expand the production capacity of K1. As for the construction of new fabs, Kioxia is staying with its plan to begin building Fab 7 in Yokkaichi and K2 in Kitakami in 1Q21. These fabs, which will be producing BiCS6 or more technologically advanced products, are expected to start contributing to the supplier’s output sometime in 2022. Technology migration will also be the main driver of its bit output growth. Currently, the majority of Kioxia’s NAND Flash products are still manufactured with the 96L BiCS4 process. Going forward, the supplier intends to raise the shipment share of 112L BiCS5 products this year.

Western Digital

Western Digital saw the ASP of its NAND Flash products drop by 9% QoQ for 4Q20 as its clients in the server segment were reducing their inventories. On the other hand, the sales of its channel-market products continued to grow, and the robust demand for notebook computers led to an impressive shipment result for its client SSDs. Western Digital’s bit shipments in 4Q20 increased by 7% QoQ. This roughly compensated for the decline in the ASP. All in all, the supplier posted a QoQ drop of just 2.1% in its revenue to US$2.034 billion.

Concerning its activities, Western Digital will be collaborating with Kioxia in the construction of Fab 7 and K2. The additional production capacity from these two plants will help the allied suppliers to deal with market competition in the future. The 96L BiCS4 process will be Western Digital’s primary technology for NAND Flash production this year. Additionally, Western Digital will be providing OEMs with samples of TLC and QLC products that are manufactured with the 112L BiCS5 process sometime between 2Q21 and 3Q21. Western Digital’s next-generation BiCS6 process is also set to enter the production stage in 2022.

SK Hynix

SK Hynix benefited from the aggressive stock-up activities of Chinese smartphone brands in 4Q20. Its bit shipments rose by 8% QoQ, but its ASP also dropped by 8% QoQ due to the general oversupply situation. With the decline in the ASP being canceled out by the increase in bit shipments, SK Hynix kept its NAND Flash revenue relatively constant for 4Q20. It posted a miniscule QoQ decline of 0.2% to US$1.639 billion.

This year, SK Hynix will rely on technology migration as the primary means of increasing bit output. The share of 128L products in its bit output came to around 30% at the end of 2020 and is expected to keep growing to surpass the output shares of 72L and 96L products in 2021. The supplier has also scheduled the launch of its 176L products for 2H21. Regarding the acquisition of Intel’s NAND Flash plant in Dalian, the transfer of the ownership of the plant along with Intel’s SSD assets is expected to be completed by the end of this year as originally planned.

Micron

Thanks to stock-up activities of smartphone brands and the growing demand for QLC SSDs from PC OEMs, Micron posted a significant QoQ increase of 17-20% in its bit shipments for 4Q20. However, like other suppliers, its ASP fell in the same period due to the general oversupply situation and registered a QoQ decline of 10-13%. In terms of revenue, Micron posted a QoQ increase of 2.9% to US$1.574 billion.

On the technology front, Micron has 128L products, but unlike other suppliers’ strategies, Micron will not ship 128L products to its main clients. Instead, Micron is focusing on the development of the second-generation 176L products that will serve as its main offerings in the future. Its clients will thus bypass the 128L process and advance directly to 176L process. OEMs are expected to begin receiving samples of 176L products from Micron in 2Q21 in accordance with the supplier’s schedule. With respect to the cell type, Micron is raising the shipment share of QLC products. Currently, more than 50% of supplier’s NVMe SSD shipments (in bit terms) are QLC products.

Intel

Intel made a recovery in its bit shipments in 4Q20 after inventory adjustments in the data center and enterprise server segments had caused a QoQ decline of nearly 25% in 3Q20. The procurement momentum of its clients was still fairly weak in 4Q20, but it did pick up somewhat compared with the previous quarter. Additionally, the demand from PC OEMs continued to rise. Consequently, Intel’s bit shipments grew by nearly 25% QoQ for 4Q20. Again like other suppliers, Intel saw its ASP drop in 4Q20 because of the general oversupply situation. The QoQ decline came to almost 20%. On balance, Intel’s NAND Flash revenue went up by 4.8% to US$1.208 billion for 4Q20.

Intel will probably not make any significant changes to its existing plans for production capacity and product mix as it has sealed the deal to sell its NAND Flash business to SK Hynix. It will continue to leverage its advantage in the enterprise SSD market to push its clients to adopt 144L products. Regarding the distribution of its product shipments by technology, Intel will be raising shipment share for the 144L stacking process and the QLC architecture. To increase the output of 144L products, Intel will expand the production capacity of the Dalian plant this year. From a long-term perspective, SK Hynix will be the main beneficiary of this capacity expansion effort.

On the whole, TrendForce’s investigations find that PC OEMs have been releasing a substantial amount of upside orders since the start of 1Q21. Although the oversupply situation is worse compared with 4Q20, it has become more moderate than expected. Moreover, the market is anticipating that clients in the data center segment will reinitiate large-scale procurement in 2Q21. The sentiment has thus turned more positive with respect to contract negotiations, and the general price decline has also begun to ease earlier than expected. Suppliers’ sales performances are projected to rebound rapidly 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

As various TV manufacturers such as Samsung, LG, and TCL announced their new models equipped with Mini LED backlights at CES 2021, TrendForce’s 2021 Mini LED New Backlight Display Trend Analysis report shows that total Mini LED chip revenue from Mini LED backlight TVs to potentially reach US$270 million in 2021, as manufacturers gradually overcome technological bottlenecks and lower their overall manufacturing costs, according to TrendForce’s latest investigations.

Mini LED backlight TVs possess a highly cost-effective competitive advantage, as Mini LED backlight costs for the entry-level segment are only 50% higher than traditional LCD equivalents

TrendForce further indicates that, with regards to TV backlight technologies, the cost of Mini LED solutions is about two to three times lower than that of white OLED and entry-level, direct-lit LCD solutions. This cost difference therefore serves as Mini LED technology’s competitive advantage over its competitors in display backlight adoption. At the moment, high-end TVs contain about 16,000 Mini LED chips per TV, divided into 2,000 local dimming zones.

In this market segment, PM (passive matrix) Mini LED TV panels with BLU (backlight unit) still cost about 15% less than OLED TV panels and therefore hold a cost advantage. On the other hand, in the mid-range TV segment, each TV contains about 10,000-12,000 Mini LED chips and 500 local dimming zones, meaning the cost of Mini LED backlight integration in this market segment is about a mere 50% more than entry-level, direct-lit LCD backlight units, making Mini LED a viable alternative to traditional LCD solutions in this segment too. Given the high cost-effectiveness of Mini LED backlight units, TV manufacturers are therefore likely to adopt them as a viable technology and initiate an industry-wide competition over Mini LED TV specs this year.

HDR and 8K resolution will be the two mainstream features of high-end TVs this year. With regards to Korean brands, Samsung’s Neo QLED Mini LED TV and LG’s QNED Mini LED TV, both unveiled at CES this year, are equipped with Mini LED backlights as a performance-enhancing technical feature. These TVs feature not only 8K resolution, but also Mini LED backlight units, which require more than 20,000 Mini LED chips (divided across more than 1,000 local dimming zones, with more than 1,000 nits in peak brightness), in addition to passive matrix FALD technology, which allows for contrast ratios of 1,000,000:1, a significant improvement that puts these TVs on almost equal footing with OLED TVs in terms of image quality. At the same time, China-based TCL is also set to release its OD Zero Mini LED TV, which has comparable specs with Korean offerings and is also equipped with Mini LED backlight units. Going forward, more and more TV manufacturers, such as Hisense and Xiaomi, are expected to participate in the burgeoning Mini LED backlight TV market.

The pace of optimizing Mini LED chips, backplanes, and driver ICs will be key to the Mini LED industry’s rapid expansion

As various manufacturers successively release their Mini LED backlight TVs this year, related companies in the supply chain are expected to benefit as a result. Currently, there are multiple major suppliers of Mini LED components on the market: Chip suppliers include Taiwanese (Epistar and Lextar), Chinese (San’an and HC SemiTek), and Korean (Seoul Semiconductor) companies. Testing and sorting companies include FitTech, Saultech, and YTEC. SMT companies include Taiwan-based Lextar and China-based Hongli Zhihui. Driver IC suppliers include Taiwanese (Macroblock, Elan, Parade, Himax, and Novatek) and Chinese (Chipone) companies. Backplane suppliers include Tawanese (Apex and Zhen Ding Tech) and Korean (Young Poong Group) companies. Panel suppliers include SDC, LGD, AUO, Innolux, BOE, and CSOT.

TrendForce believes that Mini LED backlight displays currently possess a competitive advantage over OLED displays due to the former’s 15% comparatively lower cost. Ultimately, the future development and profitability of the Mini LED backlight market in the long run will depend on the continued optimization of components that account for a relatively higher allocation of backlight costs, including Mini LED chips, Mini LED backplanes, and driver ICs.

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 the impact of the COVID-19 pandemic, global smartphone production reached a mere 1.25 billion units in 2020, a record-breaking 11% YoY decrease, according to TrendForce’s latest investigations. The top six smartphone brands ranked by production volume for 2020, in order, are Samsung, Apple, Huawei, Xiaomi, OPPO, and Vivo. The most glaring change from the previous year is Huawei’s market share.

TrendForce indicates that Honor will formally separate from Huawei and operate as an independent smartphone maker at the start of 2021. The aim behind this spin-off is to ensure the survival of Honor, which has become a major brand in the global smartphone market after years of labor. However, it remains to be seen whether the “new” Honor can capture consumers’ attention without the support from Huawei. Also, Huawei and the new Honor will be directly competing against each other in the future, especially if the former is somehow freed from the U.S. trade sanctions at a later time. With the new Honor seeking to ramp up production, Huawei will have more difficulty in regaining market share for smartphones.

Looking ahead to the rest of 2021, TrendForce believes that the global smartphone market will gradually recover as people become accustomed to the “new normal” resulting from the pandemic. Moreover, this year will likely see a relatively strong wave of device replacement demand as well as demand growth in the emerging markets. Assuming that these conditions will materialize, the annual global smartphone production for 2021 is forecasted to increase by 9% to 1.36 billion units. Regarding the annual global ranking of smartphone brands for 2021, Huawei will experience a further and significant decline in its device production. This is because of the effects of the U.S. export restrictions and the spin-off of Honor as a separate entity operating in the smartphone market. Huawei is currently projected to tumble from third place in 2020 to seventh place in 2021. The top six for 2021, in order, will be Samsung, Apple, Xiaomi, OPPO, Vivo, and Transsion. Together, they will account for almost 80% of the global smartphone market. Nevertheless, the pandemic will remain the central variable (or the biggest uncertainty) in the production projection because it will continue to exert significant influence on the global economy. Besides the pandemic, the performance of smartphone brands during 2021 could also be affected by geopolitical instabilities and the lack of available production capacity in the semiconductor foundry market.

Penetration rate of 5G smartphones is likely to rise to 37% in 2021, while production will still be constrained by limited foundry capacities

Thanks to the Chinese government’s aggressive push for 5G commercialization in 2020, global 5G smartphone production for the year reached about 240 million units, a 19% penetration rate, with Chinese brands accounting for almost a 60% market share. While 5G will remain a major topic in the smartphone market this year, various countries will also resume their 5G infrastructure build-out, and mobile processor manufacturers will continue to release entry-level and mid-range 5G chips. As such, the penetration rate of 5G smartphones is expected to undergo a rapid increase to 37% in 2021, for a yearly production of about 500 million units.

It should be noted that, under the optimistic assumption that the pandemic can be resolved within the year, shipment for various end-products, including servers, smartphones, and notebook computers, will undergo a YoY increase compared to 2020. Case in point, the number of PMICs and CIS (CMOS image sensors) contained per handset will each double in order to meet increased smartphone specifications. On the other hand, major Chinese foundry SMIC has recently been added to the Entity List once again. This is expected to exacerbate the foundry industry’s already-strained production capacity.

TrendForce indicates that smartphone brands’ recent bullish outlook towards the 2021 market and their attempt to secure more semiconductor supplies by increasing their smartphone production targets can potentially lead these brands to overbook certain components at foundries. However, smartphone brands may adjust their component inventories from 2Q21 to 3Q21 and reduce their semiconductor procurement activities if actual sales performances fall short of expectations, or if component bottlenecks remain unresolved, leading to a widening inventory gap between bottlenecked and non-bottlenecked parts. Even so, TrendForce still forecasts an above-90% capacity utilization rate for foundries in 2021.

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