LG Display (LGD) has released an earnings reports that is disheartening, at least it is if you work for LGD and not for Samsung Display or BOE. But LGD is also experiencing serious teething problems with its OLED panel plant in Guangzhou according to a late-October report in etnews | Korea IT News that references a number of anonymous sources.
It seems that LG Display applied a variety of technologies to maximize productivity in the new fab, but it wound up bringing too many new technologies on line simultaneously, which produced an unstable production system. The question now is whether LGD can stabilize production by the end of the year. The Guangzhou plant began initial operations at the end of August, and, according to “the industry,” the yield rate has not achieved LGD’s initial expectations. “[The] current yield is between 50 and 60%,” said an industry representative, etnews reported.
The Guangzhou plant is Gen 8.5, is designed to produce primarily high-res OLED panels in sizes from 55 to 77 inches. LGD intended to start production at 60,000 input panels per month, increasing to 90,000 panels and final target yield in 2021. These aggressive goals were believed feasible, said one of etnews’ industry sources, by building the plant along the same lines as the Gen 8 Paju plant that has been producing OLED panels without problems. But LGD sabotaged its own strategy by applying a number of new technologies in the Guangzhou plant that had never been attemped at Paju or anywhere else.
“The industry believes,” said etnews, that one contribution to the plant’s disappointing yield is that LGD decided to use a new supplier for some of its OLED materials. And “it is heard” that the company also changed its WOLED stack structure to increase luminous efficiency and productivity at the same time.
In a previous life, I was a semiconductor device design engineer. I was ordered to serve as the liason between our design group and the factory that made our stuff, primarily because I was the only member of our group who could talk to the factory engineers without screaming at them or being screamed at. But if I had brought the factory a design that involved changing materials and structure — and therefore process — and simultaneously required a rapid improvement in yield, they certainly would have screamed at me, and with good reason.
LGD was also planning to use multi-modal glass (MMG) — a process that allows displays of two or more different sizes to be fabricated on the same substrate — at Guangzhou. “[It] is heard” that the company will delay implementing MMG because of an initially unstable yield rate.
So how is LGD addressing Guangzhou’s unstable yield rate? Etnews says the company will implement the proven production technologies used at Paju, so the production lines at both plants will be almost identical. Wait. Wasn’t that the original strategy?
Meanwhile, on Samsung Display’s side of the fence, the grass is greener. According to Samsung’s latest earnings call, display revenues for Q3 were up 19% Q/Q at $7.75 billion, with OLED display revenues up 26% Q/Q at $6.5 billion. DSCC attributes the grown to strong demand for flexible OLEDs from Apple and Samsung and high utilization at its rigid-OLED fabs thanks to strong demand from Samsung and Chinese brands.
We wrote previously of the existence of convincing but not quite conclusive reports that Samsung Display would be investing heavily in fabs for the manufacturing of QD-OLED displays for television. (To refresh your memory, QD-OLED uses a layer of deep blue OLED emissive material and patterned quantum dots to convert the blue light to green and red for the full set of blue, green, and red sub-pixels.) Samsung Display has now formally announced its decision to invest $10.85 billion over six years in QD-OLED R&D and production lines. The first step, which begins this year, will be the conversion of an LCD line in Samsung Display’s L8 fab in Tangjong, Korea to QD-OLED production.
The Tangjong line will be converted from a production capacity of 125,000 LCD substrates per month to 30,000 QD-OLED substrates per month. with volume production starting in Q1’21. Long term, Samsung Display plans to convert all of its Gen 8 LCD lines, which will exchange 360,000 LCD substrates per month for 100,000 QD-OLED substrates per month. For this to make any sense at all, Samsung Display must convince itself that a QD-OLED display must yield more than four times the margin of an LCD. Is that realistic? Well, in Q3’19, according to DSCC, the operating margin for LCDs was -22%, while the operating margin for OLEDs was +19%. The difference is not always that stark. In Q4’18, for instance, the numbers were +6% for LCD and +12%. Overall, though, Samsung’s bet doesn’t seem to be a bad one, assuming manufacturing yields meet expectations. However, Samsung is giving itself a realistic time to iron out those problems.
One thing I don’t understand is why Samsung Display thinks it’s a good idea to base a display on the least efficient of the OLED materials. Although this is a fundamental problem, there are also positives in the QD-OLED approach. DSCC observes that Samsung will use three sub-pixels and two emitting layers compared to LG’s current WRGB architecture, which uses four subpixels and a matrix color filter. As a result, Samsung’s stack will incorporate 13 layers compared to LGD’s 22 layers, which means fewer deposition stages, improved yields, and lower material costs. In fact, DSCC estimates that QD-OLED production will have a material cost of $26 per square meter, compared to $95 for WOLED. Please note that Samsung’s QD-OLED architecture also uses a matrix color filter to absorb the small amount of blue light that passes through the areas of blue and green quantum dots without being absorbed.
(Bob O’Brien of DSCC pointed this out as a comment to a previous article
“The figures of $26 and $95 are for the OLED Stack materials only. QD OLED will certainly have a less expensive OLED stack and less complexity in the deposition process, but the cost of the Quantum Dot Color Converter may overwhelm that advantage. Perhaps more important than both of these for the cost equation will be the yields on the backplane”. – Editor)
Isn’t display manufacturing fun? (KW)
Ken Werner is Principal of Nutmeg Consultants, specializing in the display industry, manufacturing, technology, and applications, including mobile devices, automotive, and television. He consults for attorneys, investment analysts, and companies re-positioning themselves within the display industry or using displays in their products. He is the 2017 recipient of the Society for Information Display’s Lewis and Beatrice Winner Award. You can reach him at [email protected] or www.nutmegconsultants.com.