​Apple (AAPL) announced the latest iteration of the iPhone SE, the company’s answer to a low-priced iPhone version.  The last iPhone SE was released in 2020, which shortens the lag between SE models to two years, down from the earlier 4 year interval, so perhaps Apple has decided to release a new SE version again in 2024.  Physically there is little difference between the old and new models, with only a small difference in weight, and both have the same 4.7” LCD display, however the chipset has been upgraded from the A13 (7nm+) to the current A15 (5nm) which will allow for improved image processing and higher frame-rate video.  Also added is support for 5G sub6 bands and what Apple calls “the toughest glass in a smartphone”, likely Corning’s (GLW) Gorilla Glass DX, but details on the battery, which was a bit of an issue for the 2020 iPhone SE have been less detailed, although Apple indicates that the new SE should give  up to 10 hours of streaming video playback and 50 hours of audio playback, up from the previous model.  The price of the base version (64GB Unlocked) is $429, just a bit over the $399 original price of the 2020 model.
While the iPhone SE has only two cameras (12 MP & 7MP), we wonder if there is data on how often the average user intentionally picks a more sophisticated camera on the flagship iPhone models before snapping a picture.  We assume that there are some iPhone users that might need the three or four different cameras on the more expensive models for semi-commercial photography, but if shots posted on social media are any indication, most photos are taken with whatever camera is the default.  That brings us to both the need for the multiple camera trend and the idea that all the bells and whistles that are added to flagship phones seem a bit unnecessary.  Since industry leading smartphone brands tend to offer a wide variety of mid-priced phones, which are usually their best sellers, does the SE’s popularity indicate an industry feature overkill?
As the average smartphone buyer leans toward smartphone status over smartphone utility, we understand the need for flagship smartphones, especially when viewed from the brand’s margin expectations, but it seems that the popularity of the SE or the Samsung A12, Samsung’s best-selling phone which sells for $180 (32GB, unlocked), would indicate that perhaps more time should be spent reducing the price of smartphone generally instead of loading them with features that are rarely used.  We know it is antithetical to the smartphone brand mindset to think that less is more, but the SE points in that direction once again.

The fear of burn-in for OLED displays can be either a reality or a promotional tool for proponents of other display technologies, however OLED display designers have been working toward eliminating or reducing the problem since OLED display came to market.  Burn-in is not new.  It was an issue for CRT (Cathode Ray Tubes), particularly black and white displays or the monochrome monitors of the early days of personal computers where screen phosphors that were illuminated by a logo or similar image left ‘ghosts’ that were visible when the screen images were darker.  LCD displays, while there are ways in which they can also be ‘ghosted’, are less susceptible to burn-in and when it does occur, it is easily reversed in most cases.
When it comes to OLED displays burn-in was a particularly worrisome issue years ago but has been addressed in a number of ways by OLED display designers.  The cause of OLED burn-in is the uneven use of the three OLED materials that are used in each OLED display pixel (RGB displays, such as smartphones and tablets).  If an image remains on the screen for an extended time in a single location and is a particular color or shade, such as deep blue, the blue sub-pixels will be ‘on’ more than the other colors and therefore will ‘age’ faster than the others.  As OLED materials see reduced output as they age, a white screen (which implies all colors are ‘on’) where a blue image has been persistent, will see less blue in the white where the aging was most prevalent, meaning the pixels that have been persistently on will look slightly different than the other white pixels, creating a ‘ghost’ of the persistent image.
There are ways in which this can be avoided, with the most effective being not to watch programming that has persistent images, but that is certainly not a practical solution.  Second would be not to leave your display on if you are not watching it, or turning down the brightness, but there is little a user can do if they are watching CNN or sports channels where logos are extremely persistent so display designers have come up with solutions that help to mitigate the problem.  One such is to shift pixels.  At regular intervals, each pixel in an image is shifted left, right, up, or down, so the wear on a particular sub-pixel color does not fall on that pixel at all times.  This movement is not noticeable to the user but ‘spreads the wealth’ across more pixels, reducing the wear on any single one.  Extending the lifetimes of OLED materials is also a viable solution to burn-in, as the longer it takes for a material to degrade, the longer it will take for burn-in to occur under the circumstances described above, but that is a task left to material scientists.
While many studies have been done concerning burn-in, most are done in labs with display metrics measured with sophisticated equipment that would generate terabytes of data and innumerable graphs and charts, but do not accurately give a consumer a ‘real world’ picture of what to expect with OLED burn-n on a practical basis and this is where the ‘average’ Joe comes in.  The Nintendo (7974.JP) Switch offers an OLED display for its game device.  We believe the display is produced by Samsung Display (pvt), the leader in small panel OLED displays, so the quality should be above reproach, but who is willing to leave their game device on for hundreds or thousands of hours to see if it burns in?  If hundreds are willing to eat Ghost Peepers or dance naked on YouTube (GOOG), then there are those that would be willing to forego playing games for150 days to rack up a 3,600 hour ‘real world’ test to see the actual effects of burn-in.
While this is certainly not a scientific test and the conclusion is subjective, but YouTuber and Twitch (AMZN) host Bob Wulff, was willing to put his Switch through a 3,600 hour test with a single image from “Zelda, Breath of the Wild”, set to maximum brightness.  During the 1st 1,800 hours, there were no image issues but after the next 1,800 continuous hours he noticed small ‘ghosting’ images that were sometimes discernable during game play, but were easier to see when the Switch image was set to a single color full screen image.  We note that the Switch, in portable mode, does not use pixel shifting or screen dimming, so the average OLED display would have access to these fixes while the test did not.  Mr. Wulff is hoping to keep the test active until he brings the system to “an unplayable state,” and will report back with results when that happens, but for now it seems that the fear of burn-in for OLED, in this non-scientific but practical test, is rather small, but we note that the displays used in the test were produced last year, so one might be careful in researching what year a potential OLED device was produced as older display might not have the same characteristics.  See the actual video below.

https://youtu.be/PaC5RbGAeVo
​

​The Russian invasion of Ukraine presents a difficult problem for many CE companies as the Russian market for CE products is big enough that companies don’t want to burn bridges but are under pressure both from the US government and from consumers to suspend or end operations in the country.  Most recently Samsung Electronics has announced it will suspend shipments to Russia but has not indicated whether it will close its stores and Apple (AAPL) has stopped all product sales to Russia, closed its stores, and limited Apple services, however Xiaomi (1810.HK) and a number of other Chinese CE firms have yet to make a decision as to whether to continue to serve Russian customers.  Dell (DELL), HP (HPE), Google (GOOG), Apple (AAPL), Intel (INTC), AMD (AMD), Oracle (ORCL), SAP (SAP.GR), Amazon (AMZN), Meta (FB), Netflix (NFLX), Roku (ROKU), Taiwan Semi (TSM) and Microsoft (MSFT) have all suspended operations or shipments to Russia and we expect many others to join the list this week.
While the Russian smartphone market is relatively small at ~31m units, with Samsung (#1 share) and Apple (#3 share) out of the way, Xiaomi could move into 1st place if they remain.  That said, is it worth the risk of being sanctioned by the US given they use parts made in the US or by US equipment.  Probably not but most Chinese companies do not seem to be rushing to make such decisions quickly and those that are already on US trade restrictions lists, such as Huawei (pvt), ZTE (000063.CH), and SMIC (688981.CH), will likely keep their Russian relationships intact.
The list of US and foreign CE companies that have suspended Russian operations continues to grow but there are many that are slow to take action and seem more concerned as to how any operational changes in Russia would affect their bottom line.  Japanese glass supplier Asahi (5201.JP) made the following statement: “AGC Inc. hereby expresses that we are deeply concerned about the situation in Ukraine and hope for peace to be restored as quickly as possible.  The impact of the situation on AGC Group and status of its business operations in Europe and Russia are as follows.  We will continue to monitor the situation closely and take appropriate actions.”  This all seems to be a ‘if things start to affect our business, we will let you know”, rather than a condemnation of Russia’s overt acts, surprising from a company whose glass appears in many CE devices shipped to the US, risking social backlash or violations of US trade policy.  That said, there are still many US and foreign CE companies that have yet to make a statement as to their stance on Russian relations, but the continuation of hostilities over the weekend will likely push many CE companies to limit, suspend, or end their relationships in Russia.

​While Chinese panel producers continue to add Gen 8 and Gen 10 LCD capacity, Taiwan’s two large panel producers have differing opinions as to how to meet that competition.  Both are relatively conservative, with both AU Optronics (2409.TT) and Innolux (3481.TT) not having added new or additional fab capacity since 2018.  They differ however in their forward plans as AUO is adding capacity to its L6 line in Kunshan, and has announced plans to build a new Gen 8 LCD fab at a later date, which we expect will be in 2025, while Innolux has taken another path and decided not to build any new LCD fabs, but to modify its older fabs to produce higher margin products such as fingerprint sensors and mini-LED backlights.
Innolux has been using its F1 – TPO (2002) Gen 3.5 fab to produce capacitive fingerprint sensors and is expecting to expand that production to biometric identification sensors for credit cards and has announced a collaboration with Super-C Touch (pvt) and Starttek Engineering (pvt) to develop Innolux sensors for the Indian market in 2023.  India has plans to have 80% of its 1.2b inhabitants register their fingerprints over the next five years in order to boost the mobile payment industry.  Innolux says it has also converted another Gen 3.5 LCD line toward the production of Mini-LED backlights and plans to release its first 75” LCD panels using its own Mini-LED backlights next quarter.

​Last week we noted that NVIDIA (NVDA) had been hacked by a South American group known as LAPSUS$.  The hackers stole over 1TB of data including details about NIDIA’s LHR (Light Hash Rate) system that limits the performance of NVIIDIA’s graphic card processors when its graphic cards are being used for crypto-mining rather than gaming.  Graphics cards are in short supply and prices have risen as cryptocurrency miners soak up whatever stock is available leaving gamers, NVIDIA’s long-term customers, with very high prices for the cards they need and little stock.  The LHR system makes it less profitable for miners to use the cards, in theory leaving more for NVIDIAs traditional customers.  By revealing how the LHR works, the hackers are threatening to thwart NVIDIA’s system, unless they pay.
The same hacker group seems to now have hacked Samsung Electronics (005930.KS) over the weekend and mow claim to have the source code for Samsung’s Knox security system and biometric codes for Samsung’s Knox security system that resides on almost every Samsung smartphone. Tablet, or PC.  Indications are that the hackers were able to access confidential source code and other classified data, with the hackers teasing an image of a portion of the data which included the algorithms to unlock biometric data and the code behind Samsung services, which, if true, would open up millions of Samsung smartphones to the hackers or those that decide to use the hacked information.  As of yesterday Samsung was still trying to evaluate what was taken and the risk to users and has not offered a workaround or fix, but with ~20% of all smartphones sold being a Samsung Galaxy device, the problem is potentially a large one.

We have written about foldable smartphones and other OLED devices numerous times over the last few years.  Now with 7 brands having foldable smartphones, the concept is no longer new to consumers and according to recent data, Samsung’s foldable sales are reaching levels that make the concept a more realistic mass production smartphone rather than a novelty.  Vivo (pvt) a sister company to China’s Oppo (pvt) and RealMe (pvt), all three of whom are owned by BBK Electronics (pvt), is expected to join the ‘foldables’ category next month with the Vivo X Fold.
The device is expected to have a form factor similar to the Samsung Galaxy Z Fold (horizontal fold) with the main screen inside (when closed) and a large external display.  According to the most recent speculation the main display  (foldable) is being supplied by Samsung Display (pvt) the leader in the foldable display space, and the external display is expected to be supplied by China’s BOE (200725.CH), although the size of each remains unknown, and the chipset is likely Qualcomm’s (QCOM) Snapdragon 8 Gen 1 or similar.  There is a chance that the fingerprint sensor will be mounted under the display, while the camera set-up is expected to be a single selfie cam and four cameras (50, 48, 12, 8MP) for photography, all powered by a 4,600 mAh battery.
The only problem with such a device is that Vivo’s phones are most popular in China or Southeast Asia, which means that the odds of this phone winding up in the US are very small, and while foldables are a new price category for most smartphone brands, Vivo tends to offer phones in the low to mid-priced category, so we expect such a Vivo foldable will have to find an audience that is different from its usual while facing competition from Samsung in regions where it has a relatively small presence, a task that is a challenge for all foldable brands.  While we have seen data as to Vivo’s share in a number of countries as high as 36.7%, those numbers must be adjusted to the population of the countries involved and then to the number of active smartphone users.  Based on just population alone Vivo sells the vast majority (over 80%) of phones sold outside of China to India and Indonesia where the two best-selling smartphones are the Xiaomi Redmi 9 Power and the Xiaomi Redmi Note 10 Pro which sell for ~$211 and $290 respectively so a foldable in the $900+ price range (expected) will have limited appeal to Vivo’s existing customer base outside of China.
With a foldable expected from Google this year and the potential for an Apple foldable in 2023/2024, the foldable smartphone is looking to find its way into the mass market smartphone business, especially given the potential for premium pricing, something every smartphone brand is looking for in such a competitive market.  Hopefully many of the technology issues that caused problems for consumers in the early days of foldables are now completely or almost completely solved, but we expect that by the time Apple is ready to enter the foldable market, the novelty will have worn off with consumers and the objective will be to find foldables that solve consumer problems rather than fold nicely. 
Triple fold smartphones (see our note 8/31/21) give users a large image space for viewing movies or playing games, while flip type foldables bring phone sizes down to something that fits anywhere, but we always come back to a smartphone with a scrolling display, one that can be pulled out of the side of the phone to become a 6” by 12” or even 18” screen that would mimic a newspaper or magazine.  There are lots of challenges in producing such a device (How do you keep the display flexible enough to be coiled but rigid enough to stay open?) and many that we have yet to think off, but while consumers have become conditioned to watching things on a small screen, foldable should be able to change that to something more akin to a display that doesn’t force you to lean forward to catch important scene movement or force cinematographers to envision what a set or exterior shot will look like on a 6” diagonal screen.  Technology is supposed to be used to solve problems or at the least to make life easier, which gives foldable designers, who are barely constrained by historic convention, the ability to experiment as the technology develops.  Hopefully it doesn’t wind up in the same rut as regular smartphones where more ‘stuff’ means higher prices until consumers find little value in multiple cameras or more pixels and start the inevitable price decline cycle.  The field is open for foldables but using it as just another gimmick would be a true waste..  

​China’s largest LED producer Sanan (600703.CH) has been a supplier to Apple going back to 2020 but as Apple adopted Mini-LED backlighting for the 12.9” iPad Pro, the MacBook Pro, and the XDR Pro display, the specifications changed and Sanan was not included as a supplier, leaving Epistar (3714.TT) and OSRAM AMS.SW).  There have been questions as to whether the exclusion was due to product or Sanan’s IP status, but after being excluded early this year it looks like Sanan is now back in Apple’s Mini-LED good graces.  With Apple expected to add Mini-LED backlighting to a smaller iPad Pro, a 27” iMac Pro, and possibly another monitor, it was incumbent for Sanan to resolve any issues before final design stages are set.

On 2/23 NVIDIA (NVDA) discovered a flaw that could potentially allow entry into the company’s proprietary servers and began examining the issue, but three days later it was discovered that the system had been hacked by a South American group known as LAPSUS$.  NVIDIA originally thought they had neutralized the hackers and the data that had been stolen however the hackers began posting some of the information collected to prove that they still had the information.  While over 1TB was stolen that included employee information, the most important data included in the hack was details relating to NIDIA’s LHR (Light Hash Rate) system that limits the performance of NVIIDIA’s graphic card processors. 
“Why would a company known for their high processing power graphics cards want to limit their performance?, you ask, however the answer is quite simple, they don’t want their graphics cards to be used to mine cryptocurrencies, specifically Etherium mining.  The LHR system can identify when it is being used for crypto-mining and lowers the maximum hash rate of the card while the power consumption remains the same.  This makes mining less profitable and discourages miners from purchasing said graphics card for their activities.  “Why would a company not want to sell graphics cards?” you ask, but its not that they don’t want to sell cards, it is who they want to sell them to, and the answer to that question is gamers.
As we have previously noted graphics cards are in short supply and prices have risen as cryptocurrency miners soak up whatever stock is available with little regard for price as their cost is spread across a long production period.  That said, operational costs are a very sensitive subject for miners as mining is a very power intensive process, making even small changes in ongoing operating costs the difference between mining profitably or losing money.  By limiting card performance for miners but not for gamers NVIDIA expects more cards to fall into the hands of its traditional buyers, the gaming community, who have been starved for product or priced out of the market.  When LHR was first introduced it took only a few days for the code to be broken by miners but subsequent versions have been more closely tied to hardware and have continued to work against mining operations, but part of what was stolen from NVIDIA by LAPSUS$ was confidential data concerning LHS technology and the hackers are asking for $1m (in cryptocurrency) from the company or will release the confidential information concerning LHR to the highest bidder.
Fears that the hack was part of Russia’s plans for the attack on Ukraine due to its timing were erased when the hacking organization stated that they have absolutely no political affiliation (what else would a politically aligned hacker say?) and NVIDIA says the hack would have no effect on its business but given the amount of data stolen, including source code, we take both what the hacker’s said and what the company said with an ample grain of salt.  As the company has not responded to the hacker’s request for payment and the US government has made it clear that US companies should not negotiate with ransomware requests, we expect more data to be leaked and to eventually see more basic ways to defeat NVIDIAs LHR system posted.  We note that competitor AMD (AMD) does not limit GPU performance for cryptocurrency mining but under new Etherium mining rules likely to be implemented in June, Etherium mining will have less of an environmental impact but will also generate between 20% and 35% less revenue for miners, so the whole NVIDIA hack and LHR controversy could become a moot point.  Again, anyone for the gold standard?  It was so much easier to understand.

Having a single supplier of a key component is a dangerous position to be in any industry, but when it comes to OLED displays it is even more dangerous.  The most important process step in the production of OLED RGB displays, the type used in almost 750m small panel OLED displays last year, is the deposition of OLED materials which need to be precisely patterned across the display.  With pixel densities averaging almost 400 pixels/inch in recent smartphones, and each pixel being composed of at least three sub-pixels, the placement of each ‘dot’ of OLED materials is done by depositing materials through a FMM (Fine Metal Mask), a thin sheet of INVAR, a nickel-iron alloy that is very stable during temperature changes.  The mask can be thought of as a screen, although the accuracy of the space between the ‘holes’ has to be <±15 µm and the position accuracy is <± 2µm.
The masks (called ‘sticks’) are arranged in a frame that that covers half of a substrate, usually 4.9’ x 6.1’ (29.9 ft2), which enters the deposition chamber.  The OLED materials are heated in a vacuum until they vaporize and are pushed into the chamber and settle on the substrate wherever the FMM holes occur.  Given that the temperature inside the chamber is similar to the temperature needed to vaporize the materials (as high as 500⁰F), the mask material must be able to remain rigid, however while the OLED material is deposited through the mask holes onto the substrate, it also clings to the chamber walls and the mask itself.  This means the chamber must be cleaned periodically to avoid collected materials falling on the substrate and the mask must be replaced after ~1,000 uses.  As a 15,000 sheet/month OLED line runs ~500 sheets/day (1,000 half sheets), such a line would use ~30 masks each month at a cost of between $50,000 and $125,000/month/line depending on the mask’s precision.
More important than the cost of the masks is the fact that there are few manufacturers that can produce such masks at the tolerances needed, with the industry being dominated by Japan’s Dai Nippon (7912.JP), the sole supplier to Samsung Display (pvt), the largest small panel OLED producer, which leaves SDC beholden to Dai Nippon’s material sourcing, production, and pricing.  There are a few other suppliers, such as Toppan (7911.JP) and Darwin Precision (6120.TT), but they tend to deal with Chinese OLED producers and do not meet SDC’s quality standards.  It seems that SDC has decided to evaluate FMMs from a local South Korean firm, Poongwon Precision (371950.KS) to see if their masks can qualify as an alternative to Dai Nippon’s product.  Not only will the quality have to be matched but the price will have to be low enough that SDC will risk its relationship with Dai Nippon by adding a 2nd supplier.
While such qualification processes take time, we expect that even with commercial level qualification SDC will limit a new FMM supplier to a single line or fab in order to make sure the product stays consistent, while using that leverage to negotiate price with Dai Nippon.  Given that the mask business is one with high barriers to entry and Poongwon has been producing same for roughly 5 years, there are few alternatives for SDC if this relationship does not pan out and the company will remain constrained by a single supplier relationship for all of its small panel OLED production.