​Trade press in South Korea has indicated that pricing for LG Electronics’ (066570.KS) new round of OLED TVs is going to see some substantial price drops on their initial release.  Right now expectations are that LGE is going to lower the initial price of some of its new models in South Korea in order to stimulate sales and better compete with new sets from Samsung (005930.KS).  LG has three OLED TV ‘series’, with an additional ‘B’ series expected later this year, with a slightly higher price than the entry-level ‘A’ series.  While each series has its own set sizes and feature set, in general the ‘G’ series is the top of the OLED line, the ‘C’ series is the middle, and the ‘A’ series is the entry level. 
While prices have yet to be officially released, when looking at the 65” models in each series, the expected price drops are between 17.9% and 19.2%, based on last year’s models initial pricing in South Korea, while the feature set is in most instances comparable or better.  Given that 55” and 65” TVs are the most popular among consumers, we expect a similar price drop for 55” sets.  Larger models will likely see lesser price reductions, as they remain in a ‘premium’ category, but we expect LGE has been able to see the production cost of OLED TV panels decline as their fab in Guangzhou ramped production late last year.
While we also expect new 48” models to see a lesser percentage reduction than 65” models, there is the possibility that the 48” ‘A’ series entry level 4K OLED TV could be offered at $999 in the US, which would make it the lowest priced OLED TV available, considerably lower in price than the 48” ‘C’ series model, which is $1,499.  We note also that price for OLED TVs are higher in almost all regions (including South Korea) than they are in the US, so price reduction percentages in the US could vary a bit from other countries.

Just as Prince Akeem came to America from Zamunda in 1988, Samsung (005930.KS) is bringing its micro-LED TVs to the US next month and will be adding some ‘smaller’ models later this year. These sets, which were released in South Korea late last year are based on micro-LED technology, using ultra-small LEDs as red, green, and blue self-emitting sub-pixels to generate a high brightness display with almost 25m such LEDs in a 4K display.  Because of the vast number of LEDs that must be moved from die to substrate and their extremely small size, conventional pick-and-place technology will not be cost effective for mass production, and the sets Samsung is producing, even using experimental new transfer technologies, are extremely expensive, with the flagship 110” model selling for ~$156,000, quite a bit for a home TV (See table below) no matter what measure you use.  The good news is that Samsung will also be offering a 98” version to help consumers save money (?).
There are two features of these TVs that we believe set them apart, other than the price.  The first is the lack of a backlight, similar to OLED technology.  Given that the number of LEDs in typical TV backlights, even mini0LED TV backlights, is far less than the number of pixels, there are instances where sharp lines between dark and light areas in an image are lit by a single LED diode.  If the set’s image processor chooses the ‘dark’ portion of the segment, it will dim that LED, which will also dim the ‘light’ portion of the image.  If it chooses the ‘light’ portion, it will leave the LED on and ‘bleed’ light into the dark area, making it gray instead of black.  As each sub-pixel is individually controllable in a micro-LED TV there is no bleed, no halos, and no grays where blacks should be, a big improvement over LCD technology in the same way that OLED self-emissive technology has improved display contrast.  Again cost is a mitigating item for micro-LED currently, but display producers are working toward reducing production costs to make it competitive with OLED and LCD.
The other feature that we have mentioned previously is the ability of the sets to ‘manage’ the screen layout.   This allows the user to divide the screen into up to 4 segments, which in its evenly divided state, would allow for 4 simultaneous 4K programs to be viewed at one time, with sound for each individually controllable.  Whether this means you can watch four games at once, take a video call on screen without switching off other images, or reading e-mails while watching cat videos, it is an interesting feature that goes toward making the TV more of a central device rather than a dedicated receiver.  In theory this feature should be applicable to any self-emissive display with a sophisticated image processor and substantial screen memory.
Samsung has eliminated the bezels on these sets so SBR (screen-to-body ratio) is 99.99% and the sound system incorporates a system that can track an object as it moves across the screen and locate the objects sound in its 5.1 channel speaker system, but if you are going to spend enough to buy two Alfa Romero 4C Spiders, you are going to want to also buy a sound bar, like the one Samsung also announced yesterday.  This system, Samsung’s first 11.1.4 package consists of a main sound bar with 21 speakers, a wireless sub-woofer and two wireless ‘surround’ speakers, and samples the acoustics of the room via microphones in the TV, re-calibrate the sound system to fit your room’s characteristics.  While no price was given for the sound bar, a similar Samsung 2020 9.1.4 sound bar sells for ~$1,900..
While the rest of Samsung’s 2021 TV line was based on increased size choices and a few extra features, we give Samsung credit for promoting micro-LED technology at this early stage.  We expect only a small number of units to be sold and limited demo locations, but as the technology improves over the next few years we expect Samsung will be instrumental in developing a micro-LED supply chain for itself and potentially for outside customers, similar to the way LG Display (LPL) and LG Electronics (066570.KS) created an OLED TV infrastructure to support its own brand and over a dozen external OLED brands.  How long it takes for micro-LED process technology to improve enough to make it cost competitive remains an open question, but if Samsung is willing to stick their neck out at this early stage, we expect considerable sums are being spent to develop the technology. 

All three of our 5G primary indicators saw greater than 5% growth in February, with 5G form factors up 31.3% y/y, 5G vendors up 46.9% y/y, and total 5G devices up 148.2% y/y.  One category standout was 5G laptops, which had been stable at 9 models since September of last year.  In January 2 new 5G models were introduced and three more in February, and while that pales in comparison to the 306 models of 5G smartphones, we look at ‘non-mobile devices’ meaning those that would be less apt to be used while moving, as an indicator of how much 5G is being used by stationary customers.  Similarly CPE devices, equipment for 5G that remains at the ‘customer’, which grew 8% in February and is up 60.5% y/y is an even better indication as to how 5G is migrating to on premise applications, rather than being a fully mobile technology.

The three mobile carriers in South Korea are not friends.  They compete directly with each other and against MVNOs (Mobile Virtual Network Operators) in what is a highly competitive primarily urban (81.8% of pop.) market of over 51m, about the same population as Texas and Florida combined in an area just a bit bigger than Maine.  These carrier, SK Telecom (SKM), KT (KT), and LG Uplus (032640.KS) are well-known for ‘aggressive’ marketing tactics, ‘variable discounting’ based on the customer’s ‘look’, and a number of other questionable practices, so much so that the South Korean government has had to set strict guidelines on phone plan discounting.
That said, the mobile market in South Korea is robust with just over 70m subscribers (includes business customers and multiple plan subs) as of mid-2020, and in that mix about 8m 5G subscriber plans, which has grown to over 11.8m by the end of last year (almost 13m currently), and a government that provided ~$4b in tax support for 5G carrier rollouts last year.  Knowing the fact that South Korea began operation on its first 5G network less than 2 years ago makes it even more unusual that South Korea is considered the most advanced country for 5G (~22.7% share of subscribers), having placed ~166,000 5G base stations, or ~19% of the 4G base station total, which began installations almost 10 years ago.

​Last year however, the South Korean government was not keen on one aspect of 5G, mmWave, and while the spectrum has been allocated, a government study indicated that the cost of deploying mmWave coverage, because of its limited coverage, was not conducive to a widespread network and focus should be on the mid-band sub6 spectrum, where carriers had already been focused.  Equipment providers all disagreed for obvious reasons, but it seems the carriers also disagreed, and now the government has changed its stance, forming a 28GHz band 5G Common Network consultative body with the three carriers to discuss operational plans for mmWave (28GHz.).
According to preliminary plans the three carriers have agreed to share not only mmWave base stations, but also optical cables and transmission equipment and are discussing ways in which mobile customers can use the mmWave system while roaming, regardless of the carrier with whom they have a subscription.  They have even taken it further in that they are discussing creating a super ‘hot spot’ in the Gangnam  District of Seoul that will transmit ultra-high capacity content for VR and AR, again with all three carriers sharing the equipment, and the creation of mmWave BTB districts to utilize mmWave in business and industrial situations, again all shared.  The government is also now encouraging a similar common carrier mmWave extension of its sub6 (mid-band) network expansion into the South Korean suburbs.
While China gets most of the 5G global press because of the large number of base stations deployed, its area is just slightly smaller than that of the US, or 9,524% larger than South Korea.  Chinese 5G deployment is mandated and controlled by the State Government, which is certainly a plus toward expanding the network without the necessity for near-term profitability, but in South Korea, the goal of providing a robust 5G network seems to have helped to join together the three rival companies toward a common goal while reducing costs for all three, an unusual circumstance in any free market country. 
While we expect that as the US takes dubious pride in its ‘competitive economy’, US carriers might have or maybe still should think along the same lines as South Korea, rather than duplicating coverage resources, particularly for mmWave, in order to reduce the cost of reaching large numbers of subscribers across the vast space between the east and west coasts.  While it might reduce the ‘uniqueness’ of each carrier a bit, it would serve both subscribers and shareholders, but likely would be met with bluster about ‘control’ or similar nonsense unless the government was willing to ‘incentivize’ US carriers with tax breaks to encourage them to share 5G resources.  JOHO.

​After yesterday’s note about the potential change in Visionox’s (002387.CH) controlling shareholders, it seems the company has also decided to spin off its PM OLED (Passive Matrix) business as a separate company, Qingyue (pvt).  Visionox has been producing PM OLEDs since late 2008 and is among the top producers of such displays, and as such, Kunshan Visionox Display, the former name of Qingyue has two PMOLED production lines that generally operate on an OEM basis, and are capable of producing about 2,400 m2/month. This is a relatively small amount relative to the company’s AMOLEDl production level, which we estimate to be ~61,000 m2/month currently.
As we noted yesterday, the company’s largest shareholder is an entrepreneur who has been working toward monetizing his original investment in Visionox and has been continuing the company’s pattern of spinning off various parts of the display business over the years.  We can envision a scenario where Visionox lists these spin-offs to use public capital, rather than looking to private investors who could have other liquidity issues, as we suspect yesterday’s announcement alludes to.
All in Visionox is a relatively small Chinese OLED producer, and at some point the local government, who looks to be the largest shareholder, will have to decide how to fix the company and make it profitable without subsidies.  As local governments are not usually well versed in the management of display assets, in a normal environment, we would expect Visionox to struggle against much larger competitors in China and elsewhere, however if panel prices continue to rise this year, the company could get some respite from their issues, although those fortunes could change quickly as the global display environment changes.

​Next month TCL (000100.CH) will introduce a micro-OLED based VR headset that is paired to its TCL 10+ mid-range smartphone.  The VR glasses, which by themselves sell for $1,200, will be part of the package that includes the TCL 10+ smartphone (~$360) and will sell for an expected $1,145.  That said, there could be monthly charges for software needed to run the combinations and potentially for content.
The micro-OLED VR headset, which we believe is the first commercial application of micro-LEDs for VR, is based on a Sony (SNE) micro-LED display that is both light weight (said to be 1/5 the typical VR headset weight) and high resolution, with two 1080p displays, which according to the company, give the viewer the equivalent of watching a 100’ display from a distance of 4’.  The smartphone acts as the user interface and a tough pad to control VR operation.  The package will be released in Japan and Korea, with US release dependent on FCC certification timing.
 As we have noted, others are working toward high resolution micro displays, both with OLED and with micro-LEDs.  Most OLED display producers have shown at least prototypes of high-resolution OLED displays, and while OLED micro-displays can be produced at similar and even higher resolutions, they tend toward lower brightness and shorter lifetimes vs. LED based micro-LED technology, although dual stack and other processes have helped to bring OLED micro-display capabilities to a much higher level over the last few years.  Much press has been devoted to micro-LEDs, that are both bright and have considerably longer lifetimes, with companies like Apple developing micro-LED technology at its ‘secret’ lab in Taiwan,  Chinese panel producer BOE (200725.CH) working with OLiGHTTEK Opto (state) and Kopin (KOPN), and Samsung (005930.KS) workings with Sanan Opto (600703.CH).  Vuzix (VUZI) and Foxconn (2354.TT) are also working toward the release of micro-LED VR/AR products expected later this year or early next year.   That said, micro-LEDs are complicated to fabricate and are therefore expensive to produce, so the jury remains out as to which technology will be the ultimate winner in the micro-LED race, which has just been pushed forward with the upcoming TCL release..
Should the TCL VR headset get good reviews from consumers (not from ‘influencers’ who are looking to sell merch or garner ‘favor’ from CE companies), it could give higher resolution micro-OLED development projects a shot in the arm that will push them more quickly to commercial products.  While we expect Apple to eventually release a micro-LED AR/VR product next year, we expect others, particularly Samsung, to push the limits of micro-OLED for ultra-small panel applications.  While Samsung has already released micro-LED products, they are for large panel applications thus far.
 
 
 

​Digitimes Research has increased their estimate for global smartphone shipments for the 2021 year from 1.36b to 1.4b (+2.9%) and states that it expects global smartphone shipments to grow ~50% in 1Q of this year.  While we see some reasoning behind the increase, with strong iPhone sales from Apple (AAPL), we are at a loss to understand the 50% growth in 1Q shipments, not because their 340m 1Q unit volume estimate is out of line, but because last year’s 1Q global smartphone shipments were 281.7m units, which would make this year’s y/y gain 20.7%.
 As we have noted, we use an average of a number of sources to determine our shipments, but even using the lowest outlier, the y/y increase only come to 24.8%.  As Digitimes Research is not a source for us, we cannot verify the exact number they are using for 1Q 2020, but in order to see ~50% growth y/y the 1Q 2020 Digitimes smartphone shipment estimate would have to be ~225m units, quite a bit lower than the average. We do note that 1.4b units for global smartphone shipments is within what we would call normal parameters, with some estimates already higher than 1.4b, so we only question the y/y growth, not the full year estimate.
Digitimes has also indicated that they expect 5G smartphone shipments to be over 600m units, vs. 280m last year, which is a bit aggressive given that the few full year 5G estimates we have seen are in the mid-500m range, but 600m is certainly doable if things go well.  We do note that prices for 5G smartphones, relative to 4G, are close if not the same, and as integrated 5G chipsets become more common , we expect the share of new smartphone models to continue to grow, as it has in China (57% of new models in January 2021), where 5G installations are the highest.

​Due to high demand and high utilization at its semi fabs, Samsung Electronics (005930.KS) is both increasing its semiconductor investment plans and pulling in the timeline for such expansion.  Equipment suppliers have indicated that Samsung has increased its expansion plans for its Pyeongtaek 2 fab from an additional 30,000 12” wafers to 40,000 for 1Q, and boosted the annual capacity increase at the plant from 60,000 wafers to 70,000.  With this large portion of the increased yearly investment in the 1st quarter (57%), Samsung is pushing to meet the increased demand, much of which is coming from the company itself.
Samsung is also stepping up its investment in 5nm foundry production, increasing those plans by ~54% to 43,000 wafers/month and is expected to allocate its next EUV tool to the 5nm line, rather than DRAM as originally intended.  Given there are only two suppliers at the 5nm node, Samsung has little choice but to expand at least in line with competitor Taiwan Semiconductor (TSM) or fall further behind TSM, who leads the industry as to 5nm capacity. Samsung has been using EUV on at least one layer of its latest 1znm DRAM, used in the Galaxy S21 5G smartphone, which has reduced cell size by 3.4% when compared to Micron (MU) DRAM and D/R by 8.2% vs. Samsung’s last generation (1ynm).
Last month Korean exports of memory products, leader Samsung’s homeland, grew 14.9% y/y and memory prices rose in February between 3.9% and 7.9% (DRAMeXchange), with server demand expected to increase substantially (8.4% according to the SIA) this year, after a 6.8% increase in total global industry sales last year.  Samsung seems to be pushing hard to take advantage of current demand by increasing overall capacity plans and pulling in as much as possible into the 1H of this year, while competitor SK Hynix (000660.KS) recently announced that they will be spending $4.3b US to purchase additional EUV tools through 2015, with existing EUV tools to be introduced at its $3.13b M16 plant in Icheon, South Korea.

​In August of last year we noted that a spat between the COO of Hongxin Semiconductor Manufacturing Company (pvt) and the state government over funding made it into the Chinese trade press, a place usually reserved for announcements of technology achievements or market share gains.   The Hongxin project, founded in 2017, was developed to build a 14nm 30,000 wpm fab followed by a 7nm line with the same capacity, although no timetable was given.  The project received $2.24b US in 2019 and was expected to receive an additional $1.275b last year, having hired over 100 engineers and managers from Taiwan Semi,  offering pay packages as high as 2x to 2.5x TSM’s salary and bonus. 
However, despite the initial funding, HSMC was only 10% owned by the government and 90% by a Beijing private company owned by two stakeholders that have investments in a retailer, catering companies, and a number of medical firms, with no experience in semiconductors, and whose promised investment never came.  .  In fact HSMC did purchase the only 7nm mask aligner from ASML (ASML), against US government trade mandates, although after a bit of research we discovered the tool was actually for 14nm and had never been uncrated, as it was being held as collateral for an $85.3m loan made early last year.
In November, the Chinese trade press indicated that Hongxin Semiconductor was undergoing a restructuring and a management change, with the district government taking over the 90% stake it did not own, making it a completely state-owned entity. At that time a new chief executive was also named, as the former TSM executive who the company had pulled out of retirement in 2019 to lead the company, resigned from his post in June.  The project, which had originally been slated to begin trial production in 2022-2023 seems to have been put on hold since September as capital had become an issue. 
Last week it seems that Hongxin’s senior management stated that “the company has no plans to resume work and resume production,” and requested that all employees apply for resignation between February 28 and March 5, despite the February 8 response to a question about the future of the Hongxin project on the state website, which was “This project is coordinating and advancing.”  In October of last year the Chinese Ministry of Industry & Information Technology and the National Development & Reform Commission indicated that “Enterprises are engaged in the integrated circuit industry, and individual places do not have enough understanding of the law of IC development, blindly start projects, low-level repeated construction risks appear, and even some projects have stagnated…resulting in a waste of resources,” a nice way of saying that many do not know what they are doing when it comes to developing semiconductor projects in China.
The state followed the study that came to this conclusion with new ‘specifications’ that require funds to be in place and product technology be available before employees are hired, to ‘protect and regulate the market,’ all under the seemingly more watchful eye of the state government.  Last year we discovered two other semiconductor projects in China, Decoma Semiconductor (pvt), a project sponsored by the Nanjing Economic & Technical Development Zone, and Dehuai Semiconductor (pvt) which was stalled after the $1.8b project ran out of money, despite construction being almost 90% completed (phase 1), with the local district government taking over the project and looking for funding.  In just the 1st half of last year more than 140 IC related projects were launched in China, with a total investment of 307b yuan ($46.9b US), with many local government and regional support groups looking to help China become less reliant on the global semiconductor industry.  Perhaps a bit more oversight is necessary this year.

​Wearables, particularly watches, can be helpful if you need to look at the time or date, or even the number of steps you have taken during the day, but in order to accomplish the myriad of functions included in most wearables, you need two limbs, one arm to hold the device, and a hand to press the buttons or touch sensors.  This usually means that whatever you are doing, you must stop and use both hands to perform any functions.  There are some wearables that respond to voice commands and a few that respond to gestures, but by and large, if you want to do anything with most wearables, you need both hands.
Gesture based systems have been under development for years and despite standardization, has never quite caught on, so wearable developers look for other alternatives to make it easier for watch and other wearable users to perform more than one or two basic functions or control other devices with their devices.  While we don’t usually mention products that are under development by small companies, given that many never actually see the light of day or are far from what the developers have promoted, we came across an interesting ‘accessory’ for the Apple (AAPL) Watch that allows the user to perform functions with only one hand.
The accessory is a watch band that is loaded with sensors that monitor electrical signals that are being sent to your brain as you move your fingers and through AI algorithms (of course) figures out which fingers are moving and translates those functions into control signals for your Apple Watch, which means you can control the watch, and consequently an iPhone or similar device just by moving the fingers on one hand. Typically gesture-based devices use cameras to recognize hand or facial expressions, which means the user either has to be wearing AR/VR type glasses that contain a camera.  In the case of the Mudra device, the gestures are ‘sensed’ rather than ‘seen’, allowing single handed operation.  Thus far we have seen only demos, but bypassing the need for optical recognition of any sort radically simplifies the control process and could encourage users and developers to use gestures rather than voice functions, which have their own issues.
The Mudra Band is part of an Indiegogo campaign that has raised $216k for Wearable Devices Ltd (pvt), an Israeli company that is a graduate of the Samsung Runway accelerator program and won a CES 2021 Innovation and ‘Best Wearable’ honor in January.  The company currently sells a development kit for $500 or a $3,000 Enterprise kit that allows the user to incorporate the Mudra gesture control into other products, with a tentative shipment date of March for the retail band, which sells for $179.  We note that we have no stake in this company or campaign and have an interest only from a technical standpoint.
Watch the video below to better understand how it works.

Watch the video below to better understand how it works.
youtu.be/pnY1QzwFIKE
​​cdn.idropnews.com/wp-content/uploads/2021/02/26140117/Mudra-Band-GIF.gif