The MicroLED Monitor thread

ScifiGeek

Ars Legatus Legionis
16,349
Often when discussing upcoming monitors we see a comment along the lines of: "Waiting for MicroLED". This is the thread to give some sense of how long that wait might be, by posting any actual info, and/or solid theories in the absence of information. Usually there is no Industry indicators, and we just resort to pointing out the enormous complexity. I found an industry timeline in this PDF, from the MicroLED Association, that actually mentions consumer products of interest like TV, Laptops and Desktop Monitors, and IMO this is probably optimistic, but it's a starting point:


Here is the Page with the Timeline. In Short, there is a projection to have "Pilot Level Production" in 2030-2033.

The odd one for me, is that they seem to have a slight lead for Monitors, when I would Expect TVs to be first as you can scale down the tiled approach to large consumer TV's and those can be very expensive already. So I expect we will see MicroLED TV's before Monitors:

MicroLED.PNG
 

BO(V)BZ

Ars Tribunus Militum
2,082
That's somewhat depressing to read, but thanks for the information. After hearing about them in the media for years, just to learn there's a decade or so to wait seems almost like a placeholder for "we have no idea" IMO. I don't trust the tech industry to plot hardly anything more than 5 out, but either way it's clear we aren't going to get 5000nits screens with a backlight for every subpixel any time soon.

I wonder, are they basing this info on trends for screen resolution as well? I expect by 2030 virtually all TVs will be 8k - not that they need that high of a pixel count, but just to get people to upgrade. There likely will be high-end 16k screens as well.
 

sakete

Ars Scholae Palatinae
826
Subscriptor++
I think that by the time MicroLED is viable for commercial production in normal sized displays, OLED tech will have advanced enough where longevity is no longer a real concern (it'll practically last 5-7 yrs without burn-in and without needing to baby it). And it should by then also be significantly more affordable than microLED.

One near-term advancement in oled tech should be pholed, where the organic material used in the blue oleds will go from fluorescent to phosphorescent, which allegedly will greatly improve durability (I.e. significantly reduce the decay rate).

But like the zen master said, we'll see.
 

redleader

Ars Legatus Legionis
35,019
I think that by the time MicroLED is viable for commercial production in normal sized displays, OLED tech will have advanced enough where longevity is no longer a real concern (it'll practically last 5-7 yrs without burn-in and without needing to baby it). And it should by then also be significantly more affordable than microLED.
Conventional MicroLED (where you put millions of individually manufactured LEDs onto a circuit board) is never going to be viable for commercial production outside of insanely expensive wall displays. There will have to be some advance that enables manufacturing the LEDs in situ rather than making them separately and then gathering millions of them together. For example, the self-emissive quantum dots linked above would work, since you can potentially make them through deposition. So would improvements in organic emitters (also made by deposition).

Really the question isn't when will you be able to afford a MicroLED TV (answer: when you win the lottery) but when will someone figure out a way to build pixels in situ that have comparable properties to ones made lithographically.
 

cogwheel

Ars Tribunus Angusticlavius
6,691
Subscriptor
The odd one for me, is that they seem to have a slight lead for Monitors, when I would Expect TVs to be first as you can scale down the tiled approach to large consumer TV's and those can be very expensive already. So I expect we will see MicroLED TV's before Monitors:
The part that stands out to me is tablets/laptops/monitors before smartphones. Tablet & (smaller) laptop screens aren't much lower pixel density than smartphones, so the smaller overall display size and often somewhat smaller resolution would seem to favor smartphones first.

One of the Ars members who is in a position to see the early screens says they are having burn in issues with microLED as well, and that it does not greatly improve that aspect of the display over OLED, at least right now.
I wonder what is going on here. The whole point of microLED is that the LEDs use the exact same extremely durable doped semiconductor construction as normal LEDs instead of the organic electroluminescent construction (that degrades much faster) of OLEDs.
 

cerberusTI

Ars Tribunus Angusticlavius
6,449
Subscriptor++
The part that stands out to me is tablets/laptops/monitors before smartphones. Tablet & (smaller) laptop screens aren't much lower pixel density than smartphones, so the smaller overall display size and often somewhat smaller resolution would seem to favor smartphones first.


I wonder what is going on here. The whole point of microLED is that the LEDs use the exact same extremely durable doped semiconductor construction as normal LEDs instead of the organic electroluminescent construction (that degrades much faster) of OLEDs.
I am not entirely sure, but there are a few ways that could happen. Displays are relatively intolerant of differential aging of colors especially.

It was a note here:

He runs https://blurbusters.com/ and does a bunch of work with various display manufactures, and is more likely to know what is going on in the display industry than most.
 

redleader

Ars Legatus Legionis
35,019
The part that stands out to me is tablets/laptops/monitors before smartphones. Tablet & (smaller) laptop screens aren't much lower pixel density than smartphones, so the smaller overall display size and often somewhat smaller resolution would seem to favor smartphones first.
A 4K display has the same number of pixels no matter the display size, so making the display smaller doesn't help with costs unless you can cut the resolution too. Essentially the bigger display just spaces the same number of pixels differently.
 

cogwheel

Ars Tribunus Angusticlavius
6,691
Subscriptor
I am not entirely sure, but there are a few ways that could happen. Displays are relatively intolerant of differential aging of colors especially.

It was a note here:

He runs https://blurbusters.com/ and does a bunch of work with various display manufactures, and is more likely to know what is going on in the display industry than most.
I'm talking about microLED, which is entirely different from OLED despite both ending in "oled". OLED organic emissive layer failure should have nothing to do with microLED failures, since microLED doesn't have the organic stuff at all.

A microLED display should be using small but otherwise conventional red, green, and blue LEDs, so it shouldn't have the phosphor degradation issue white LEDs have since it doesn't use that phosphor layer at all.

Your link also doesn't go to a specific comment, just the fifth page of comments on that story, so I don't know what comment you were trying to link to. You might need to link to the specific comment using the forum view of the comments.

A 4K display has the same number of pixels no matter the display size, so making the display smaller doesn't help with costs unless you can cut the resolution too. Essentially the bigger display just spaces the same number of pixels differently.
Sure, but smartphone screens rarely are 4K. For example, a Galaxy S23 Ultra has slightly more than half the pixels of a 4K screen, and the smaller S23s (and all iPhones) have even fewer pixels.
 

cerberusTI

Ars Tribunus Angusticlavius
6,449
Subscriptor++
I'm talking about microLED, which is entirely different from OLED despite both ending in "oled". OLED organic emissive layer failure should have nothing to do with microLED failures, since microLED doesn't have the organic stuff at all.

A microLED display should be using small but otherwise conventional red, green, and blue LEDs, so it shouldn't have the phosphor degradation issue white LEDs have since it doesn't use that phosphor layer at all.

Your link also doesn't go to a specific comment, just the fifth page of comments on that story, so I don't know what comment you were trying to link to. You might need to link to the specific comment using the forum view of the comments.


Sure, but smartphone screens rarely are 4K. For example, a Galaxy S23 Ultra has slightly more than half the pixels of a 4K screen, and the smaller S23s (and all iPhones) have even fewer pixels.
It was a comment regarding microLED if you feel like searching that page, not OLED.

Try this (pasting the link apparently goes there, but also rewrites the URL to the page and not the comment):


Edit:
Anyway, there are not enough of them in the wild that I can find anything online about this, but that is one of the more likely people to have some idea as to how those first displays are doing.
 
Last edited:

cogwheel

Ars Tribunus Angusticlavius
6,691
Subscriptor
It was a comment regarding microLED if you feel like searching that page, not OLED.
I'm not saying microLED can't burn in or dim, just that if it does, it's from a fundamentally different reason from every other "LED" technology mentioned in that comment. The OLEDs dim/burn in due to their organic compounds in their emissive layers breaking down. MiniLED dims because miniLED is just a backlight technology, and the white LEDs it uses have a phosphor layer that breaks down over time from usage.

MicroLED is basically a standard LED (not OLED, not white LED), so it doesn't have organic compounds or phosphor layers to degrade. It's bulk semiconductor doped with different elements.

My guess is that if microLED is having degradation problems worse than OLED, it's probably due to heat causing electromigration or metal diffusion. That's probably at temperatures beyond what would cook OLEDs, so that implies that it's much harder to keep microLEDs cool than it is other LED families.

Try this (pasting the link apparently goes there, but also rewrites the URL to the page and not the comment):

That worked, thanks!
 

cerberusTI

Ars Tribunus Angusticlavius
6,449
Subscriptor++
I'm not saying microLED can't burn in or dim, just that if it does, it's from a fundamentally different reason from every other "LED" technology mentioned in that comment. The OLEDs dim/burn in due to their organic compounds in their emissive layers breaking down. MiniLED dims because miniLED is just a backlight technology, and the white LEDs it uses have a phosphor layer that breaks down over time from usage.

MicroLED is basically a standard LED (not OLED, not white LED), so it doesn't have organic compounds or phosphor layers to degrade. It's bulk semiconductor doped with different elements.

My guess is that if microLED is having degradation problems worse than OLED, it's probably due to heat causing electromigration or metal diffusion. That's probably at temperatures beyond what would cook OLEDs, so that implies that it's much harder to keep microLEDs cool than it is other LED families.


That worked, thanks!
A little bit of searching for the likely mechanism yields a fair number of articles in journals talking about sidewall defects and damage in LEDs at this size, where it becomes significant. Red is apparently more problematic than green or blue.

This one has an image in the abstract which may help, and is investigating the issue.
 
  • Like
Reactions: continuum

redleader

Ars Legatus Legionis
35,019
Sure, but smartphone screens rarely are 4K. For example, a Galaxy S23 Ultra has slightly more than half the pixels of a 4K screen, and the smaller S23s (and all iPhones) have even fewer pixels.
You're missing the point there. Saying we can cut costs by half on something that is 3 or 4 orders of magnitude beyond economical does not make it more "favored". It's drinking 4000 times the lethal amount of cyanide vs. 2000 times. Neither is favorable because the outcome is the same either way.

If any sort of inorganic emissive diode technology becomes economically viable it'll be based on completely different technology than lithography and so have different cost scaling anyway. It's nonsense to extrapolate from the current dead end approach.
 

ScifiGeek

Ars Legatus Legionis
16,349
Last edited:

Exordium01

Ars Praefectus
3,977
Subscriptor

Paranoid Android

Ars Scholae Palatinae
835
Subscriptor
Interesting, I just stumbled upon this thread after reading some decidedly unpromising news about what was supposed to be a breakout moment for MicroLED commercialization.


What should we make of this? Is MicroLED a technological dead-end in the foreseeable future or should we still expect to see progress being made?
 

ScifiGeek

Ars Legatus Legionis
16,349
Interesting, I just stumbled upon this thread after reading some decidedly unpromising news about what was supposed to be a breakout moment for MicroLED commercialization.


What should we make of this? Is MicroLED a technological dead-end in the foreseeable future or should we still expect to see progress being made?

At each size "MicroLED" is essentially a different technology. Success/Failure at one size doesn't really tell us much about other sizes because they are so different.

"MicroLED" at Watch size, is most likely fabbing the display right on some kind of wafer. I expected this one would be one of the more likely spaces to productize, since you aren't having to deal with ridiculous amounts of individual parts.

At the other end of the scale is the "MicroLED" we currently have. Giant TVs. Really, these are more like mini-LED backlights, without the need for a LCD in front. It isn't that big a technology challenge because it's fairly low density, which is why these exist. But it is a ridiculous amount of parts to deal with and properly connect, but that is more covered by the typical $100K+ pricing, than some kind of advanced technology.

Finally "MicroLED" at Monitor size is what I expect we may never see. It's very high density, but still has to be made of individual parts. It's both an incredible technology challenge, with no good route to affordable costs. While you can charge some $100K for a Giant TV, you won't be able to do that for a 32" Monitor.
 
Last edited:
  • Like
Reactions: continuum

Paranoid Android

Ars Scholae Palatinae
835
Subscriptor
Interesting. I was somewhat aware that there were serious manufacturing challenges holding MicroLED back, and that smaller displays in particular were especially difficult, but having it described as "essentially a different technology" and having problems "fundamental to its design" puts it rather starkly!


When OLED first started appearing in TVs about a decade ago, manufacturing difficulties were often cited as a limiting factor, and it was only in the last five years that OLED has started to really make inroads in the TV and PC monitor markets. I sort of assumed that MicroLED would follow a similar trajectory after the first wave of ultra-expensive TVs hit the market, but it sounds like that kind of straightforward development path isn't really applicable here.
 

Exordium01

Ars Praefectus
3,977
Subscriptor
At each size "MicroLED" is essentially a different technology. Success/Failure at one size doesn't really tell us much about other sizes because they are so different.

"MicroLED" at Watch size, is most likely fabbing the display right on some kind of wafer. I expected this one would be one of the more likely spaces to productize, since you aren't having to deal with ridiculous amounts of individual parts.

At the other end of the scale is the "MicroLED" we currently have. Giant TVs. Really, these are more like mini-LED backlights, without the need for a LCD in front. It isn't that big a technology challenge because it's fairly low density, which is why these exist. But it is a ridiculous amount of parts to deal with and properly connect, but that is more covered by the typical $100K+ pricing, than some kind of advanced technology.

Finally "MicroLED" at Monitor size is what I expect we may never see. It's very high density, but still has to be made of individual parts. It's both an incredible technology challenge, with no good route to affordable costs. While you can charge some $100K for a Giant TV, you won't be able to do that for a 32" Monitor.
I’d think the watch display would still involve transfer printing three separate subpixels from three separate wafers. I agree on the last point though. If costs are incurred by pixel instead of by inch, computer monitors will be impossible to sell.

So that means MiniLED is going to be what we have for a while unless you want to deal with slow response times with dual-layer IPS or with possible burn in with OLED.

I think miniLED/FALD backlights have already lost to OLED. IPS panel tech is just too leaky (backlight bloom is a serious issue on Apple’s entire product line) and Samsung has allegedly given up on VA panel development.

OLED burn-in seems like a mostly solved issue if you don’t need to compete with sunlight and can set the brightness to something reasonable along with the standard slight over provisioning with periodic pixel shifting and refreshes.

My guess is that we are either one generation away from or on the first generation of OLED computer monitors that aren’t all that susceptible to burn-in.
 
Last edited:

ScifiGeek

Ars Legatus Legionis
16,349
I’d think the watch display would still involve transfer printing three separate subpixels from three separate wafers. I agree on the last point though. If costs are incurred by pixel instead of by inch, computer monitors will be impossible to sell.

Then you have the same problem as with Micro-OLED monitors of assembling millions of individual subpixels.

I'd think you would want some kind of chip scale of full display fabbed on chip, like these guys are doing:

https://compoundsemiconductor.net/a...ng_The_Manufacture_Of_MicroLED_Micro-displays

They make the display right on the wafer. No cutting out pixels and re-assembling them. Though these guys are only at the monochrome stage. But there is work on full RGB chip displays as well, but I can't find it right now.


Unless they can fab display on wafer, I think Micro-OLED is dead on watches as well.
 

Exordium01

Ars Praefectus
3,977
Subscriptor
Then you have the same problem as with Micro-OLED monitors of assembling millions of individual subpixels.

I'd think you would want some kind of chip scale of full display fabbed on chip, like these guys are doing:

https://compoundsemiconductor.net/a...ng_The_Manufacture_Of_MicroLED_Micro-displays

They make the display right on the wafer. No cutting out pixels and re-assembling them. Though these guys are only at the monochrome stage. But there is work on full RGB chip displays as well, but I can't find it right now.


Unless they can fab display on wafer, I think Micro-OLED is dead on watches as well.
The point of transfer printing is that you aren’t picking and placing individual dice.

That article suggests combining three monochrome displays with optics... not something that will fit into a wearable. Also, that approach requires significantly more wafer area for the same result.
 

ScifiGeek

Ars Legatus Legionis
16,349
The point of transfer printing is that you aren’t picking and placing individual dice.
If you are transfer "printing" by fabbing the individual subpixels on three different types of wafers, cutting them out, placing them together in proper pixel grouping on a substrate, then using that substrate to "print" them, it's really more like an advanced pick/place technique than printing. I don't see this competing economically with OLED for watches. Watches don't seem to be a harsh use case for OLEDs and I haven't notice many complaints about burn in.

That article suggests combining three monochrome displays with optics... not something that will fit into a wearable. Also, that approach requires significantly more wafer area for the same result.

Yes, I know this isn't the solution. There are others working on this monolithic fabbing technique with full color on one chip. I just couldn't find a link. Also you could possibly just use the blue version, with some quantum dots deposited to change the blue to Green and RED for the other subpixels (which is how QD-OLED works), but that requires some serious precision to do the QD deposition on subpixels that small.
 

Exordium01

Ars Praefectus
3,977
Subscriptor
If you are transfer "printing" by fabbing the individual subpixels on three different types of wafers, cutting them out, placing them together in proper pixel grouping on a substrate, then using that substrate to "print" them, it's really more like an advanced pick/place technique than printing. I don't see this competing economically with OLED for watches. Watches don't seem to be a harsh use case for OLEDs and I haven't notice many complaints about burn in.
Transfer printing is nothing like pick and place. It's still essentially a film transfer process. There is one alignment/registration process for everything getting transferred as opposed to aligning individual units.

The problem with OLED displays on watches is that they aren't bright enough for high sunlight.
 
Transfer printing is nothing like pick and place. It's still essentially a film transfer process. There is one alignment/registration process for everything getting transferred as opposed to aligning individual units.

It's very much like it. Except that there are fewer component types (R, G, B, subpixels), and the placement is more regular, since you are laying down regular lines, instead of placing components mixed about.

The three subcomponents are fabbed separately on different wafer types, and have to be collected and assembled with great precision onto some kind of transfer strip, before you lay them down. Most likely into strips similar to the strips that feed a pick and place machine.

Sure there are differences from doing some kind of heat transfer with with the pixel strips, than dropping individual components, but the concept is extremely similar, and while it will be cheaper to transfer 8 million sub pixels with a heat transfer technique, than a traditional pick and place technique, it's still going to be expensive.

So far I don't think there are any actual microLED products that use these transfer techniques. The current "MicroLED" TVs are massive walls with individual pixels spaced out on boards, most likely using old style pick and place.
 
Last edited:

Exordium01

Ars Praefectus
3,977
Subscriptor
It's very much like it. Except that there are fewer component types (R, G, B, subpixels), and the placement is more regular, since you are laying down regular lines, instead of placing components mixed about.

The three subcomponents are fabbed separately on different wafer types, and have to be collected and assembled with great precision onto some kind of transfer strip, before you lay them down. Most likely into strips similar to the strips that feed a pick and place machine.
You are right up through the three sub components fabbed on three different wafers. What you describe after with strips isn’t how transfer printing works. It shares more similarities with a film transfer process than an individualized element transfer process.
 
Last edited:
You are right up through the three sub components dabbed on three different wafers. What you describe after with strips isn’t how transfer printing works.

They have to precision assemble the the three subpixels onto some kind of transfer film (either together, or three seperate transfer mediums applied in three steps), which is is going to be much more analogous to some of the pick and place complexity, but it comes in a different place in the process.

So, you are essentially have similar or greater complexity, you just have it at different place in the process.

Putting together a reel for pick and place, is trivial, and more of the complexity is in the placing step.

Putting together a transfer medium for microLED transfer is the and extremely precise/complex step, which simplifies the later transfer. Precision for true microLED is insane, as these are dust particle size emitters. So you need insane precision in both the creation of the transfer medium and it's final application.

There are differences in the process, but high level concept is similar, and complexity is if anything higher for true MicroLED.

And again, I don't think there is yet a single product in existence that does this kind of transfer for true MicroLED. The so called "MicroLED" product currently available, are macro scale, and use actual pick and place of larger LEDs for display walls.
 
Last edited:

Exordium01

Ars Praefectus
3,977
Subscriptor
They have to precision assemble the the three subpixels onto some kind of transfer film (either together, or three seperate transfer mediums applied in three steps), which is is going to be much more analogous to some of the pick and place complexity, but it comes in a different place in the process.

So, you are essentially have similar or greater complexity, you just have it at different place in the process.

Putting together a reel for pick and place, is trivial, and more of the complexity is in the placing step.

Putting together a transfer medium for microLED transfer is the and extremely precise/complex step, which simplifies the later transfer. Precision for true microLED is insane, as these are dust particle size emitters. So you need insane precision in both the creation of the transfer medium and it's final application.

There are differences in the process, but high level concept is similar, and complexity is if anything higher for true MicroLED.

And again, I don't think there is yet a single product in existence that does this kind of transfer for true MicroLED. The so called "MicroLED" product currently available, are macro scale, and use actual pick and place of larger LEDs for display walls.
Again, what you are describing isn’t transfer printing. You don’t assemble devices onto a transfer film. The film goes onto the devices on wafer and then you remove the substrate. You get to the pixel pitch you want by dilating the film in x and y after substrate removal.

Nobody is doing what you are describing here for MicroLED. You are confusing MicroLED and miniLED. There are no mass produced microLED products on the market.

The only product on the market that claims to be MicroLED that I’m aware of is Samsung’s $150,000 tv. And I’d imagine that they are eating multiples of that per TV sold in yield in order to be able to say that they are first to market. MicroLED is not a commercial technology right now. Yes, the transfer printing is hard. It’s one of the things keeping MicroLED from reaching the market.
 
Last edited:
Again, what you are describing isn’t transfer printing. You don’t assemble devices onto a transfer film. The film goes onto the devices on wafer and then you remove the substrate.

Then that would be an extra intermediate step. Because the pitch on a wafer is dramatically smaller than it will be on a display.

So you would then need to transfer to yet another medium with a different pitch that matches your display, or are you claiming you then use a wafer sized printer, to precisely place a mass of pixels sparsely selected individually out of the dense pack, and keep reposition, and sparsely selecting them from the remainder? That's insane.

The reality is that there isn't yet any one way to do this, and there are multiple options including even "fluidic" assembly being tried.

Apple spent 10 years and Billions on this, with watches as the the first "easy" target, and apparently recently cancelled all that massive resource spend because they determined it couldn't be done profitably.

So again, there are no products that do true MicroLED with any of these advanced transfer techniques. Those are still research stage.

All we have is essentially macroScale pick and place LED Walls.