Originally Posted by slacker711
Young claims that they have hit acceptable yields using SMS in that link.
I have never had your definitiveness on this issue. Samsung believed that they could do it and clearly failed in their first attempt. Could they have had a manufacturing breakthrough? We'll know if/when they start spending on a Gen 8 RGB fab.
Slacker, I don't actually doubt that the yields are acceptable.
What I doubt is that the throughput is acceptable. And I doubt strongly it ever will be acceptable.
My definitiveness goes to Operations 101. In order to produce a screen using mask scanning, you need to move a finite number of small masks over a larger substrate (1-4 seems plausible, with 2 being a likely maximum). Whatever the number, however, the process is inherently slow. You can improve the deposition of OLED through the screens, but it's still a mechanical process in two ways (1) the pushing of the OLED materials (2) the movement of the screens. The screens need to be move in perfect registration or else the panel / substrate is ruined.
You can get this kind of precision because, well, in clean-room type manufacturing, nearly any level of precision is possible. You cannot, however, have that level of precision while also having high throughput. The very movement of the masks has to be deliberate to keep the registration perfect. The OLED deposition has to be cautious to avoid errant material.
The decision to go with a full RGB design makes this even worse as you have to deposit each color separately.
Samsung gets away with this on mobile because, honestly (1) there is no mask scanning and (2) they have brute forced the production of mask-based OLED production to the point they are good at it. And they can afford to just toss out bad screens. And the screens don't have to be very high quality. Now, I don't mean they are low quality in terms of image reproduction because obviously the latest ones are just flat-out excellent. I do mean in terms of longevity, however. A smartphone screen needs to work for <10,000 in nearly all cases. If the MTBF is anywhere around there, you will have an acceptable defect rate. A tablet screen might need a slightly higher MTBF and, I suspect, they are being more cautious with tablet screens as a result, which is why you'll still see them sell far fewer OLED tablets than OLED smartphones (market size helps, but be aware of the mix, too.)
If you are talking about making 20-30 million large-format televisions annually down the road, this method won't scale. It's not at all clear it will scale to 2-3 million, but it's very clear it won't scale to the bigger numbers. The promise of OLED is embedded in reasonable build cost at high volumes. If your "Herbie" is a process step that is nearly as bad as early LC sputtering on large-format LCD TVs, it's a dealbreaker. It was for LCD until someone fixed the problem and it will be for any attempts by Samsung to hit volume production here.
Therefore, two possibilities exist:
1) They have no intention of hitting volume production. This is a long-term investment and they are just making sure they are in the game.
2) They have made an unannounced breakthrough that is game changing with respect to how they will produce these. Even if they have not done this, they will ultimately need to.