Smartphone chipsets and processors are designed to be highly specific to their use cases, and while they can also be used in tablets, laptops, wearables, and some smart TVs, they are designed to be the hub that controls most of the functions of your mobile devices.  There are only a few major independent smartphone processor brands, Qualcomm (QCOM), Mediatek (2454.TT), and UniSoc (pvt), but a number of smartphone brands produce their own application processors and have for many years.  These processors are either exclusive to the brand or are sold externally, but in most cases, they remain internal, with independent suppliers designing and supplying the majority of smartphone brands.
That is not to say other brands have not tried their hand at designing their own silicon, with Samsung (005930.KS) among the earliest with its Exynos 3 processor, used in the 2009 Galaxy S, and Apple’s A4 processor, used in the iPhone 4 in 2010.  2016 was a big year for self-designed application processors, with Lenovo (992.HK) releasing its Tiantian processor in its Vibe K5+ phone, and ZTE (000063.CH) releasing its N1 processor in its Axon 7 model that same year.  Huawei (pvt) joined the club with its Kirin processor series in 2020, although that was made expedient when the US banned exports to the company, and Google (GOOG) more recently with its Tensor processor series.  The remaining brands typically use Apps from Qualcomm and MediaTek, who battle each other on a model-by-model basis.
The problem with commercial application processors are designed to perform a variety of tasks well, but are not designed specifically for a particular smartphone brand or model, which tends to make smartphone models and brands more generic. For example, there are currently 32 smartphone models that use the Qualcomm Snapdragon 8 Gen 1 processor, across a broad swath of brands, from Samsung to Xiaomi (1810.HK), and 16 models use the MediaTek Dimensity 9000, including One Plus (pvt), Honor (pvt) and Vivo (pvt), while 3 models use the internally designed Samsung Exynos 2200 processor, all of which are Samsung phones.  With all three having release dates withing 3 months of each other, the commercial Aps tend to reduce the diversity between models and brands.
Then there is price, and while with so few AP producers there is likely less price competition than there should be, custom APs are considerably more expensive to produce considering the smaller number of units.  That said, as noted above, a number of smartphone brands do design and have produced their own APs.  However, as the smartphone business is facing another year of declining volume, the cost-effectiveness of such development programs is being questioned, and only 3 days ago, Oppo (pvt), the 4th largest global smartphone brand (~10% share in 4Q ’22) shuttered its chip design company ZEKU (pvt) with no advanced warning to employees, customers, or HR, who was hiring new employees up to the 11th. 
Oppo had released its first self-developed chip in 2021, along with its own power management and Bluetooth audio silicon and had been said to be taping out its own 4nm processor recently, while building a $650m chip R&D center.  The company’s CEO said that current revenue did not meet expectations and that the investment in self-developed chips was so large that the company could no longer afford to support the project and issued its termination.  That said, some employees seem to differ with those comments, noting that the AP processor tape-out is not due back from Taiwan Semi (TSM) until late Jun, and if successful, would have put considerable pressure on Qualcomm.  They take it further in that they say that company executives have been in the US earlier this month and were warned by the US that Oppo faced sanctions if it continues with its current self-developed processor path, in order to protect Qualcomm’s position.
None of these allegations are supported with documentation at this point, but it seems that a warning from the US, after the devastation that has been done to Huawei’s smartphone business, might be enough to shutter what has been a ~$7b investment for Oppo, rather than be added to the US ‘naughty’ list and see its share of the smartphone market disappear in a few months’ ala Huawei.  Shades of Teddy Roosevelt’s 1901 speech where he said, “Speak softly and carry a big stick”.  Maybe they were smart enough to realize that the economics were not good no matter what the situation was in a market that is not growing or maybe they realized that they would not be able to buy an semiconductor equipment going forward and would fall behind commercial AP manufacturers.  Too soon to tell but its not a good sign for the Chinese semiconductor industry and one that will empower US politicians to continue to keep the pressure on China.

average guy or gal in the street might not know much about this 5-day event, or even care.  In most instances there is little reason for the average Apple user to pay much attention to what is discussed at the conference, as they will likely not see the actual changes to their devices until much later in the year if they notice them at all.  But this year could be different in that not only will Apple present developers with a new version of iOS (iOS 17), and new versions of iPad OS, MacOS, Watch OS, and TV OS, as they typically do, but the company is expected to announce the long-awaited XR headset that has been rumored for years.
Apple’s XR development program has been around in its current form since 2017, but Apple has been making acquisitions relating to AR/VR since 2014 (see below) and allows developers access to a number of tools and resources that they can use to create AR and VR applications for iOS and iPad OS and have access to the Apple Store.  But Apple itself has not championed a physical XR headset, disappointing developers and fans a number of times in the past.  The most recent ‘rumored’ device, the Apple Reality Pro, would be Apple’s entry into the XR hardware space, and would be a driver for the industry, that in the long run, would likely have even more impact than Meta’s (FB) Quest series of VR headsets, as Apple’s
hardware following, with over 2 billion active devices, to Meta’s ~20m Quest headsets sold.  Of course, Meta has over 2b active Facebook users, along with Instagram, Messenger, etc., but hardware is Apple’s thing.
If Apple does decide to make an announcement concerning an XR device, it will be to stimulate developers to build or modify apps to operate under what will likely be a new sub-OS specifically designed for Apple’s XR hardware, seemingly called xrOS.  This code will work under iOS and will allow the device to communicate and process applications that reside on a secondary device, such as an iPhone or iPad.  Making the headset a ‘non-standalone’ device, reduces the bulk, weight, and computing power on the headset itself, making it more comfortable to wear, a concept not lost to Apple developers who are likely following many Apple design mantras that are consumer oriented over technology oriented, while tethering also means at least one other Apple device must be purchased or owned for the headset to be operated.
Expectations are that the headset battery, which in some headsets is in the headset itself, will likely be external, perhaps a belt or pocket clip-on, with a cable to the headset, while the communication between the headset and the paired device could be a cable or wireless, but the Apple XR device is also expected to not sport controllers, which would be a departure from the norm.  It has been suggested that Apple will use gestures to control the headset, with a large number of cameras scanning body movements, eye movements, area mapping, and even facial expressions for information that the xrOS can use to keep the user’s field of view correct.
The displays are expected to be something close to 4K micro-OLED displays, with on-board processing through Apple’s own SoC, either the M2 or a specially designed processor, with a focus on power efficiency given the battery-driven nature of the device, but while developers will be interested in the internals, consumers will be more concerned with how it looks, how it feels, and can it do things other AR/VR headsets cannot, and the idea that the device can operate as both a VR and AR device is not new.  Other VR headsets allow a black mask to be removed from the headset, allowing it to operate as an AR device through cameras, so the proof will be in the applications themselves and how easily the headset operates. 
​As we have noted recently eye-tracking is becoming  more popular as it facilitates foveated rendering, and movement tracking are common, so it will be incumbent on Apple to devise a gesture system that is intuitive for consumers, but on an overall basis, for Apple to have a successful product, especially on that is expected to cost between $2,500 and $3,000, it must be able to provide functionality, as while there will be an initial rush from ‘I always buy the newest Apple product’ people, it better provide more than ‘coolness’, and much of that will come from applications.  While the iPhone is a well-designed device, as are most Apple products, it also serves as a communication hub for its users, and while we do not expect an Apple XR headset to become as ubiquitous as an iPhone in the near future, it has to be able to do more than play games or allow you to buy things in the Metaverse.  We keep our expectations low but our hopes high.

​In or note earlier this week we mentioned that two of the three telecommunication carriers in South Korea have had their 28GHz (mmWave) band licenses revoked by the Ministry of Science for non-fulfillment of contracted obligations., and we mentioned that the 3rd carrier, SK Telecom (017670.KS) was likely to lose its 28GHz band license by the end of the year.  As it turns out on May 31 SK Telecom will be added to the group as the Ministry of Science has found that the company was only able to connect 1,650 devices to its 28Gh. Network, falling considerably short of the 15,000 it had promised almost 5 years ago.  SK Telecom responded, “After allocating the 28GHz frequency band, we have been making continuous efforts to create an ultra-high frequency band ecosystem and discover business models (BMs). I'm sorry it wasn't there”.  As South Korea has been a leader in developing other telecommunication modalities (2G/3G/4G) it is surprising that the country has been unable to continue that leadership in 5G.

apan Display (6740.JP) is a company in transition, actually a company that has been in transition for quite a while.  Japan overall had been a dominant supplier of LCD displays, starting with wrist watches in the early 1970’s, progressing to calculators, and to hand-held TVs (‘pocket TVs’) by Seiko (6724.JP) in Japan in the early 1980’s, and by the late 1980’s Japanese companies dominated the production of LCD displays and LCD based products.  As the millennium approached, South Korean and Taiwanese companies began to challenge Japan’s position in the LCD space, and in 2011 three Japanese LCD display producers, Sony (SNE), Toshiba (6502.JP), and Hitachi (6501.JP) merged their small/medium size LCD businesses together to form Japan Display, along with a $1.49b US ($1.642b today) from the Innovation Network Corporation of Japan, a partnership between the Japanese government and 19 corporations that has been loaning capital to Japanese companies since 2009.
JDI was successful in building a small/medium size LCD panel business with the capital and assets donated by each partner, along with production assets purchased from Panasonic (6752.JP) and was listed on the Tokyo Exchange in 2014, however by 2015, price competition and relatively high operating costs pushed JDI into the red, despite being the primary small panel supplier to Apple (AAPL).  Unfortunately for JDI, in 2017 Apple began a gradual change from LCD displays for the iPhone to OLED displays, and JDI’s small panel mobile business has never recovered.  A restructuring in 2017 led to substantial layoffs and the closure of the company’s Hakusan LCD fab, which was eventually sold to Sharp (), but as Apple continued its iPhone transition to OLED JDI was left to find a new customer base.  Apple itself made an investment of $100m in 2019 as the company restructured a second time, with more plant closings.  JDI was able to develop enough OLED capacity to supply Apple with watch displays but was so far behind in OLED development that it made little difference.
After mounting losses in 2018 and 2019, JDI ceded control of the company to Ichigo Asset Management (2337.JP) for$715m in capital, and at the time the company indicated that they would begin development of a proprietary OLED technology that would be commercialized by 2022.  We give considerable credit to Ichigo management, who has brought costs down far more consistently than previous managements and has made progress on the promised OLED technology known as eLeap (see notes from 2/10/23, 4/10/23, 4/27/23, 5/10/23) the uses photolithography for the placement of OLED materials rather than the more typical thermal deposition.
However as JDI moves closer to becoming profitable again as part of the overall display industry, they have been facing the same weak demand and pricing issues as other display producers, and that has slowed the process toward positive earnings further.  Citing “Triple punch of higher costs, lower demand, & lower capacity utilization driving large losses across global display industry – Rising material & energy costs, global inflation, rising interest rates, & global economic slowdown remain significant headwinds”, JDI reported fiscal 2022 (3/2023) of ¥270.7b ($2.02b US), down 8.5% y/y but slightly above the company’s forecast, while net income (loss) declined further, increasing from ¥8.1b last year to ¥25.8b this year.

On the positive side, JDI signed an MOU with HKC (pvt) (see 4/10/23 note) for the development of JDI’s eLeap OLED technology, which they believe will generate royalties in fiscal 2023, along with an additional ¥18b in reduced fixed costs, but on a net basis, operating profit this year will only improve by 9%, against a sales decline of 11.3% as the company continues to face lower volumes and rising material costs as it develops new products, with eLeap mass production beginning in 2024.  In the near-term JDI will continue to reduce its exposure to the commodity smartphone business (down 36% in f. 2022), increase its exposure to automotive displays (up 26% in f. 22) and bring back growth to their VR and wearable display business (down 15% in f. 22), which faced lower volumes as customers were forced to raise prices, all of which are the ‘standard’ year-end ‘fixes’ for most small panel LCD manufacturers.
 While it is hard to put perspective on JDI’s short-term growth strategy given the current economic state, we do give them credit for posting intended actions the company will take to return to ‘mega-growth’ by 2026, some of which are financial, most of which involve Ichigo financing.

1. Ichigo lends JDI ¥20b to repay ST loan from INCJ, which puts Ichigo loans to JDI at ¥48b.
2. Ichigo buys ¥53.7b of JDI debt from INCJ which eliminates all JDI debt to INCJ.
3. Ichigo forgives ¥15b of JDI debt.
4. INCJ returns Class A preferred shares to JDI at no cost. Shares cancelled.
5. Ichigo does ¥86.7b debt swap with JDI (@¥45) on Ichigo loans to JDI, which eliminates all JDI debt.
6. JDI issues ¥173.6b warrants to Ichigo (@¥45) which will provide capital for JDI growth.

All of which will happen by the end of fiscal 2023.  With the finances as above, the company management (Ichigo) is predicting its long-term growth as follows, and while there are so many variables out 5 years that we have little faith that the path will remain intact, we have to give them credit for putting the numbers on paper and Ichigo for making a large bet.

​Remember AR & VR?  Just a few months ago, the media was full of images of folks in VR headsets, gesturing in the ether, or wandering through animated worlds where you could be anyone you wanted.  That was until AI came along and the tech ‘concept of the week’ shifted.  While those who have nano-second attention spans and feel they can trust just about anything that is posted on social media that has a cool picture attached, are satisfied to flit from one tech topic to another on a weekly basis, most investors have a longer timeframe (at least some), particularly VCs.  Although VC investments are down this year, there are certain areas where early-stage capital is available, and while the Chinese economy has been relatively weak, it seems funding for AR/VR and some associated components is still available, perhaps in smaller does, but still available in the AR/VR space.
Here are a few that have been completed recently:
MojoVision (pvt) - $22.4m – US - MojoVision is developing high-density Micro-LED displays for AR/VR.
VueReal (pvt) - $10.5m – Canada – Micro-LED (high-density) displays.
Thunderbird Innovation (pvt) – $14.5m – China – AR headset developer backed by TCL (000100.CH)
Pimax (pvt) - $30m – China - Primary supplier of VR headsets.
Maiyun AR - $?? – China – Micro-LED design and transfer solutions.
Suzhou Lilong Semiconductor (pvt) - $15m – China – Display driver ICs and Micro-LEDs
 
 

Google (GOOG) has now (almost) joined the list of smartphone brands that have released foldable smartphones, with the announcement of the Google Pixel Fold, the company’s first foldable model.  With Samsung on its 4th yearly iteration, and foldables from 7 other major smartphone brands, Google, and Apple (AAPL) were among the few that have not joined the foldable craze.  The Pixel Fold itself, is as close as one might come to a Galaxy Z Fold 4 clone, with only fractional differences in size and features, and, not unsurprisingly, the same price as the Z Fold 4, although if you pre-order, you can get a Google watch for free.
As the Pixel Fold is still to be delivered there is little real-world data to see how consumers find the software and applications, but there is one feature that seems somewhat noteworthy.  When using the main screen to make a presentation, the presenter is able to see the same images on the back screen, giving the ability to make sure the verbiage is in sync with the display presentation.  It’s a simple application, but a helpful one that gives some indication that Google is thinking more about applications than hardware, which is likely a better direction for them to go, rather than battle Samsung point for point.
Growth expectations are considerable for foldable smartphones this year, with estimates ranging between 30% and 51% growth in unit’s y/y, but the absolute numbers are still relatively small, and even smaller looking when compared to the overall smartphone market.  Even at the top end of the growth estimates, foldable smartphone unit volume is likely to be less than 2%, but with the average smartphone selling for ~$300, foldables sell at a huge premium.  Samsung Electronics, who has been the leader in the foldable smartphone space for years, had an over 80% share last year and with such premium pricing, and a weak smartphone market, we expect more of the same this year, and likely a new model (aside from the Z Fold/Flip 5 series) for the holidays, which will reinvigorate the hardware competition between brands,  and potentially give those who have passed on foldables thus far, a reason to become more interested.

Back in the early 1900’s Sharp (6753.JP) was a metal fabricator who came to fame when the company’s founder developed the first ‘global mechanical pencil’, but saw that business destroyed in a 1923 earthquake.  Moving the business to Osaka, Sharp began producing Japan’s first radio sets, and by 1953 was producing TVs, calculators, and microwave ovens 10 years later.  Sharp had developed LCD technology for its calculator business, opening a Gen 1 line in 1991, but began investing heavily in LCD technology in the early 2000’s, building a Gen 4.5 fab in 2003, a Gen 8 fab in Kameyama in 2003, and the world’s first Gen 10 fab in Sakai, in 2009, which was able to produce, at the time, the world’s largest TV, a 65” model more efficiently than any existing LCD fabs.  Unfortunately, the global economic crisis and the strong yen around that time wreaked havoc on the cost of Japanese produced panels and CE products, and Sharp’s financials continued to deteriorate until 2012, the company’s 100th birthday year.
After months of negotiations and a major reduction in price after continued stock price declines, Foxconn (2354.TT) purchased a 66% stake in Sharp for $3.5b (original price was ~$6.25b), while the Chairman of Foxconn took a 50% personal stake in the Sakai Gen 10 fab[1], with the capital expected to go toward maintaining the LCD business and funding Sharp’s development of OLED displays.  As Sharp had been an early adopter of IGZO (Indium Gallium Zinc Oxide), a backplane technology used in large OLED devices, the plan was sound, but never fully materialized, although the company was the first to produce a commercial 8K TV as early as 4Q ’15, a dubious honor currently, but under the guidance of Foxconn, Sharp became profitable in 2017.
Sharp was not an exception to the display industry’s woes in 2022, and just reported its first loss since the Foxconn takeover (for the year ending in March ’23).  While sales were up 2.1% y/y, Sharp reported an operating loss, with 84.6% of the loss a write-down of assets ($1.64b), of which 85.4% was against the company’s LCD business, 9.6% against the OLED business, and 4.9% against Sharp’s other businesses. The company indicated that lower selling prices, leading to lower sales, FOREX, and the heavy losses in the display business, were the reasons for the loss, while mix and cost reductions provided some offset, and while management emphasized that all segments except display saw sales increases in fiscal 2022, only one division saw a positive operating profit.
Company guidance for fiscal 2023 (3/24) was a bit darker than other CE companies stating “We expect the demand environment to remain weak in response to a reactionary decrease in demand from the COVID-19 pandemic, global inflation, high energy costs, geopolitical issues, and other factors.  At the same time, carbon-neutral, DX, and other sectors will remain strong.  We are seeing an easing related to the impact of semiconductor shortages, high raw material prices, and rising logistics costs; however, the future remains uncertain.”  The company cited a 2023 goal of achieving net profit at all costs, through cost reduction, and the building of the company’s brand businesses, predicting a 0.5% increase in y/y sales and a modest profit (0.4% margin).  This assumes a 34% improvement in display profitability, a 32.7% improvement in Sharp’s brand[2] business profit, and a small (+2.9%) improvement in the company’s device[3] business operating profitability.  Inventory levels across the company and more specifically in display have fallen since their recent peak in 2022, but against an average level during the three years before COVID (2017 – 2019) of ¥226.3b and an Inv./Sales ratio of 1.19, there is still some trimming that needs to be done before the company could be considered ‘lean’.
While the Foxconn management has done a good job of removing the blinders that were on the pre-sale Sharp management, panel producers overall were unable to avoid the effects of both a return to pre-pandemic demand and catastrophic inflation.  While the company did not break out the Sakai Gen 10 plant from others, we expect the large panel LCD TV business to see a recovery in 1H as prices for TV panels have gained ground, but we also agree with management’s assessment of demand in a post-pandemic environment, and now that inventory has been returned to more normal levels, demand will be the key to more than a ‘back-to-baseline’ year.


[1] Stake was sold back to Sharp in 2022 for $296m by Gou, making it part of consolidated income.

[2] Brand Business includes appliances, energy solutions, TV, and mobile devices.
 

[3] Device business includes display, IoT, and semiconductors.

​As we have noted previously, AI systems are not just rows of processor cards with blinking LEDs but are systems that must be almost spoon-fed information upon which algorithms provided to them, make decisions when queried.  Before such systems become viable, they need to be taught, using vast amounts of data and ‘teaching’ algorithms that are provided by data engineers and are separate from processing algorithms.  That said, before the systems are taught, the data must be ‘organized’ into data sets that the system can recognize, which must be accomplished by humans sitting in front of computers.  Each of these processes are different, involving both humans and computers and computers alone, but they all have one thing in common and that is that they all use energy.
While the current AI focus has put such systems in the public eye, but computers have been sucking down power for years, and while models can make estimates of what percentage of the global power grid they use and how much it increases each year, there is no specific monitoring, or is there a way to come up with an accurate estimate.  The models estimated that in 2018 ~1% to 2% of the global power supply was used by computers, with 2020 as high as 4% to 6%, and models predict that at that rate it will reach anywhere from 8% to 20% by 20230, and most of that modeling was done before the recent AI fanfare and funding.
Carbon emissions are certainly a concern as demand increases but there is also a longer-term concern that with that kind of consumption growth, capacity will not meet demand.  AI systems will get more efficient and consumption/cpu will decline, but if CPU volume growth continues or accelerates, as it seems to be current doing, who will get ‘power priority’, the average TV viewer or the AI system that is running the local telecommunication system?  Of course, this is a bit sensational, although still a concern when trying to balance a green environment and continued demand for energy, but Microsoft (MSFT) seems to be thinking ahead a has signed a deal to ‘guarantee’ its potential energy needs in the future.
According to reports, Helion (pvt) a start-up funded by Sam Altman, the current CEO of OpenAI (pvt), the company that developed ChatGPT, Dustin Moskowitz, a founding member of Meta (FB), and Reid Hoffman, the Co-founder of Linkedin (MSFT), has agreed to provide Microsoft with power it generates by 2028.  While this might not seem unusual, Microsoft likely has an army of negotiators working with utilities across the globe, this source is a bit different as Helion’s technology is based on nuclear fusion.  Helion is so positive that they can accomplish this goal that they have agreed to generate at least 50 Megawatts in the following year or pay a penalty, all of this with the understanding that to date fusion technology has not produced any electricity.    Helion has even signed an agreement with Constellation Energy (CEG), the owner of the largest nuclear power plant in the US, to manage the Helion project and market its output.
Helion has promised to demonstrate a working prototype next year, with Mr. Altman adding, “Our goal is to make the coolest tech demo in the world.  Our goal is to power the world and do it extremely cheaply.”  Of course such optimism comes after one pumps $375m into Helion but he insists Helion will be able to deliver power to the global grid by 2028 and checks in on the company roughly once a month.  The fusion industry did receive a boost last December when scientists at Lawrence Livermore National Laboratory were able to generate more energy that they put in during a fusion reaction, but on a net basis that is a far cry from being a commercial fusion producer at a cost that is competitive with other energy modalities.  Is Microsoft drinking Altman’s very expensive Kool-Aid after signing a multi-billion-dollar partnership with Altman’s OpenAI, or do they know something no one else does?  We should know by the end of next year.

OLED displays are exceptional in that they have infinite contrast, the ratio between the blackest black and the whitest white.  They typically have a wide viewing angle, averaging ~140⁰, while LCD displays have a more narrow viewing angle of ~80⁰, meaning the contrast is reduced less as you move away from the center of the screen (Figure 1).  The colors tend to be ‘saturated’, meaning OLED materials have narrow color ‘peaks’ (Figure 2).  Phosphorescent OLED materials are more efficient than LCD displays that need a backlight, they are thin and flexible, and OLED materials have a very rapid response time (typ. 0.01ms for OLED vs. 1 – 16ms for LCD).

​OLED & LCD Color Filter Spectral Comparison - Source: Transmission Spectra – Journal of Display Technology

Based on the above, every display used in consumer devices should be an OLED display, providing the best possible image to customers, but there are issues, the largest of which is cost.  While in theory, OLED displays require no backlight as they are self-emitting, and for smaller OLED displays, do not require a color filter, both of which are necessary for LCD displays, the BOM should be lower for OLED display products, but that is not the case.  In a typical OLED fab, OLED materials are vaporized in a heated chamber and pass through a Fine Metal Mask, essentially a very fine screen, that ‘places’ the OLED material in precise positions to form an RGB pixel on a substrate.  Masks must be very thin and very rigid in order to avoid ‘shadows’ that will cause the materials to be misplaced, but as more holes are added to produce higher resolution displays, the masks become more expensive and subject to gravity, causing misplaced pixels, leading to low display yields and an overall higher display cost.

​Slot & Slit Type FMM detail - Source: Toppan

Over the years display equipment engineers have come up with some alternatives to mask-oriented OLED deposition, particularly ink-jet printing, and while there have been a relatively small number of IJP OLED displays made available commercially, the process, in terms of directly printing OLED materials, has its own limitations.  In order for OLED materials to be printed, they need to be dissolved in a solvent, and that can change the properties of some OLED materials, and as the materials must remain in liquid form in order not to clog the ~50,000 nozzles that are small enough to drop 6 pico-liters of material for each sub-pixel (a pico-liter is equal to 1 trillionth of a liter), with a typical 4K RGB OLED display requiring 24,883,200 droplets.   In high resolution displays, where the pixel density is high, a drop that is even a bit too large can cause the material to migrate into another pixel, which leaves IJP to some of the less precise deposition layers, such as encapsulation materials, rather than OLED material deposition itself.
There is an alternative process that is currently being developed by a number of display producers that is based on photolithography, rather than mask-based deposition.  Japan Display (6740.JP) has been developing a maskless photolithography process called eLEAP (Environment positive Lithography with Maskless Deposition, Extreme long-life, low power, & high luminance, Any shape Patterning) that promises 2x the brightness of mask-based deposition displays, 3x current display lifetimes, while reducing 150,000 tons of CO2 emissions/year.  JDI has plans to commercialize the process by 2025 in partnership with China’s HKC (248.HK), and rumors that Samsung Display (pvt) has decided to test JDI’s process. 
A typical (not that there is a typical process) OLED display being processed without a mask would go through the following steps:

1. Clean the substrate.
2. Coat the substrate with OLED material (one color)
3. Apply resist and cure.
4. Pattern vis plasma etch.
5. Strip resist
6. Repeat steps 2 – 5 for each color.

With semiconductor photolithography stepper tools, theoretical line width down to 1um could be patterned, which would lead the way to higher resolutions that would prove extremely challenging for mask-based deposition, but there are drawbacks, a number of which need to be solved before the process becomes scalable or cost effective.  As the process indicated above uses an open-mask or sputtering system, the cost/m2 should be a bit lower than a FMM system, but deposition tools are only produced by two or three manufacturers and can cost upwards of $100m, depending on size and complexity.  Add to that the cost of an i-line or DUV stepper, reticles, and photomasks, and you have added anywhere from $25m to $120m to the start-up cost of such a line, all of which goes into the panel cost.
There are other issues that need to be addressed, particularly the potential effect of the resist on what are typically very sensitive OLED materials, and the effects of UV curing radiation.  There are also questions concerning how the plasma etch process itself might compromise the integrity of the material stacks and a host of other questions that would influence the commercialization of such a process.  So it comes down to physics and chemistry, and then a hefty dose of process engineering to make this concept into a viable display that can compete with other OLED deposition methods, and other existing or potential display modalities.

​One point that will help is that display manufacturers are at least familiar with the photolithography process, as they produce the thin-film transistor backplanes that drive all displays, but those processes are well-known and mature, while photolithographic deposition is relatively new.  The good news is that this month Chinese OLED panel producer Visionox (002387.CH) announced that it is introducing ViP (short for Visionox Intelligent Pixelization), what the company says is the world’s first metal-mask free RGB self-alignment pixelation technology, and while that ‘first’ may be contested by JDI, the Visionox process claims to be able to increase the brightness of a ViP display 4 times over metal-mask OLEDs, increase the device life by 6x, increase the light-emitting area from 39% (mask) to 69%, and increase the pixel density to over 1,700 ppi, with flagship smartphone displays at between 400ppi and 500ppi.
Visionox has indicated that it has produced medium sized samples based on the technology and is ‘rapidly advancing the work related to mass production’, which, when completed will be applied to AR/VR, wearables, phones and even TVs, and has even begun to build a ViP batch production line in Hefei.  While there is still much uncertainty regarding the development and production timeline, the fact that two panel producers are seriously considering the technology is likely to expedite R&D efforts and possibly overcome some of the existing obstacles, and then it will become a direct cost issue if it is to become a practical and profitable process for high-resolution displays.  A few years to go, but the fact that it would not require building a new display infrastructure certainly gives one hope that photolithographic deposition can join the other display processes and technologies that are in operation or on the horizon over the next few years.

The US ban on semiconductor equipment being sold to Chinese fabs has a significant political component, as there are few public figures who would take the stance that China does not pose a threat to the US way of life, and while that does not justify the bans in themselves, China’s generous funding climate for the semiconductor space gives fuel to the idea that “…if we don’t stop them, they will take over the semiconductor market”, despite the fact that until recently the US government has woefully underfunded the US semiconductor production effort, giving rise to the current semiconductor issues with China.
There is some justification for the semiconductor industry’s progressive move to China over the last decade, but primarily for the semiconductor packaging segment, which is far more labor intensive than chip production, but the real driving force is that China is a huge market that needs to be fed and years ago did not have the capacity to produce what it needed.  In fact China is still (2022) a net importer of semiconductors, and saw its semiconductor trade gap increase last year to $261.7b, increasing 46.7% y/y, so despite all the hoopla over China’s potential takeover of the semiconductor space, they remain a large net importer.
That said, the downside to the US semiconductor equipment bans to China have both an obvious financial downside for those US companies that have supplied such tools to Chinese fabs in the past, and for companies that do so, but are not situated in the US, which brings up the point that the US stance must be adopted by many countries and companies that do business with Chinese semiconductor fabs and design houses, as to violate the US rules could cause political or diplomatic friction and potentially cause unwanted repercussions for same. 
But in the case of semiconductor trade rules, the rules are not always the rules for everyone, and companies like Samsung Electronics (005930.KS) and SK Hynix (000660.KS), that are major semiconductor manufacturers and have fabs in China, are looking for ways around the US rules in order to stay competitive.  According to sources in Korea, the US Department of Commerce has been in discussion with the South Korean government concerning the aforementioned restrictions, in order to come up with a ‘separate’ solution for Korean companies that have fabs on the mainland.  The US has already given Samsung and Hynix a one-year moratorium on the ban, which would have applied to both company’s fabs in China, but that agreement runs out in September, which would mean that both companies would be unable to upgrade facilities in China, which would have serious implications for both.
As the restrictions specify tools capable of producing at 16um or lower, DRAM at 18um or less, and NAND at 128 layers or more, this would put the South Korean companies at a major disadvantage as Chinese NAND players move up to 200 layers, as would be the case with DRAM, where China is already producing at 17um, and at 14um on a more general basis.  Without the equipment necessary to stay ahead of Chinese fabs, both companies would have spent billions on capacity that will eventually far behind local producers.
The separate rules are being discussed because South Korean companies are a large part of the semiconductor market, and tacitly to maintain and develop the semiconductor production in the US by those same companies, so exceptions are being developed to potentially allow the Korean companies to import US developed tools that are able to produce a level above what is being produced in China.  While this seems necessary in terms of the potential political and financial chaos it would cause if relations between the US and South Korea were to sour, it does open the door to others, particularly Taiwan Semiconductor (TSM), asking for the same or similar exceptions for its China fabs.  The US has already convinced the Dutch to join the ban, as Netherland-based ASML (ASML) is the world’s largest supplier of DUV and EUV lithography tools, although those restrictions are still pending, and little is known as to what the US might have promised to come to an understanding with the Dutch.  How fast can the US tap dance?